Validation of a Novel Cost Effective Easy to Produce and Durable In Vitro Model for Kidney-Puncture and Percutaneous Nephrolitholapaxy-Simulation

Abstract

Introduction:

Nephrolithiasis is one of the most common diseases in urology. According to the EAU Guidelines, a percutaneous nephrolitholapaxy (PNL) is recommended when treating a kidney stone >2 cm. Nowadays, PNL is performed even for smaller stones (<1 cm) using miniaturized instruments. The most challenging part of any PNL is the puncture of the planned site. PNL-novice surgeons need to practice this step in a safe environment with an ideal training model. We developed and evaluated a new, easy to produce, in vitro model for the training of the freehand puncture of the kidney.

Materials and Methods:

Porcine kidneys with ureters were embedded in ballistic gel. Food coloring and preservative agent were added. We used the standard imaging modalities of X-ray and ultrasound to validate the training model. An additional new technique, the iPAD-guided puncture, was evaluated. Five novices and three experts conducted 12 punctures for each imaging technique. Puncture time, radiation dose, and number of attempts to a successful puncture were measured. Mann–Whitney-U, Kruskal–Wallis, and U-Tests were used for statistical analyses.

Results:

The sonography-guided puncture is slightly but not significantly faster than the fluoroscopy-guided puncture and the iPAD-assisted puncture. Similarly, the most experienced surgeon's time for a successful puncture was slightly less than that of the residents, and the experienced surgeons needed the least attempts to perform a successful puncture. In terms of radiation exposure, the residents had a significant reduction of radiation exposure compared to the experienced surgeons.

Conclusion:

The newly developed ballistic gel kidney-puncture model is a good training tool for a variety of kidney-puncture techniques, with good content, construct, and face validity.

Introduction

In patients with kidney stones >2 cm, percutaneous nephrolitholapaxy (PNL) is the recommended surgical treatment.¹ Due to technical advances in minimally invasive percutaneous surgery, a shift toward endourological percutaneous procedures has taken place in recent years.² Now kidney stones <10 mm are managed by percutaneous surgery.³

The crucial aspect is the precise puncture of the caliceal system.^4,5 Often, it is necessary to perform a freehand puncture without needle guidance. This technically difficult aspect must be trained in a safe environment under ideal conditions. There are ethical concerns of training a novice surgeon during real surgeries.⁶ As a result, the role of PNL-simulators embedded in a clinical curriculum has been strengthened.^6,7 Though the need is obvious, there is still a lack of completely validated training models. We developed and evaluated a new, easy to produce, in vitro model for the ideal training of the freehand puncturing process of the kidney.

Materials and Methods

The following criteria are optimal parameters for our training model. It must be easy to produce, feasible for all imaging modalities (fluoroscopy, ultrasonography, CT, and MRI), of human-like consistency, cost effective, easy to transport, and durable.

In the literature describing existing kidney puncturing models, one limitation was an inability to display the sonographic puncture tract. Our idea was to invent a pig kidneys/aorta model embedded in ballistic gel to overcome this problem.

Ballistic gel is a special kind of gelatine often used in forensic and criminal science to investigate the effect of different weapons, mainly firearms, on the human body. This material allows visualization of the tract of a projectile or other physical disruption. In addition, different materials, for example, animal bones or organs, can be embedded in the media.

We used a prepared pig kidney/aorta organ model (Optimist, Innsbruck, Austria) consisting of a pair of porcine kidneys with adherent ureters (about 3–4 cm length), including parts of the retroperitoneal fat and the vascular structures: aorta and renal arteries, renal veins, and a part of the inferior vena cava.

Both ureters were intubated with 5F ureter catheters (IMP, Karlsruhe, Germany) and fixed with sutures. Additional sutures were placed at the superior and inferior poles and lateral convexity of the kidney. The holding sutures should be at least 8 cm long.

We used TM Ballistic Gel Type A 300 Bloom (TMP Products, Bad Camberg, Germany) to prepare the gelatine. According to the recommendations of the company, 3 L water was heated to 50°C and 800 g of granulated gelatine was added with continual stirring. After the granules were fully dissolved, the solution was allowed to cool for 2 hours until set. The solid material was again reduced to small pieces and dissolved in 1 L water in a standard cooking pot (minimum 5 L capacity). The temperature was raised to 60°C until the mass was reduced to a liquid state. To mask the kidney/aorta model in the transparent gel-surrounding, we added a mixture of red and blue food coloring (2 × 4 g Brauns Heitmann food coloring powder) to get an opaque media, so the kidneys are completely invisible (Fig. 1).

FIG. 1.

Kidney/aorta model in ballistic gel with ureter catheters. Fluoroscopy-guided puncture.

If a CT scan of the model was planned, contrast agent Urolux Retro^® 150 mg/mL (Sanochemia Diagnostics Deutschland GmbH, Neuss, Germany), was added after the food coloring powder. Foam bubbles that occurred on the surface of the liquid gel were dispersed with defoaming agents.

Finally, the aorta-kidney packet was placed in a plastic box (Lock & Lock, Frischhaltebox HPL880 4.6 L; Lock&Lock Co. Ltd., Frankfurt, Germany), in the anatomically correct position, with the ureters pointing downward. The holding sutures were fixed outside the box using regular fixing plaster.

The heated, liquid gel was poured onto the organ model until the box was 60% filled. The position of the model can be optimized by manipulating the holding sutures. After 30 minutes the rest of the gel was poured onto the kidney packet to cover the entire organ packet. The box was closed with the refastenable lid and placed into a refrigerator for at least 3 hours until the gel has cured.

The full validation process of a training model includes seven different validation parameters. Face validity describes the extent to which the examination resembles the situation in the real world. Content and construct validity define the extent to which the intended content domain is being measured by the assessment exercise and to which extent the model differentiates between a good and a bad performer or groups of performers. Concurrent validity characterizes the extent to which the results of the test correlate with known gold standard tests and the predictive validity describes to which this assessment will predict future performance. To improve the learning strategy on behalf of the trainer and the trainees are shown in the test results as educational impact. The last parameter is cost effectiveness, which shows whether the model provides maximum value for money.⁸

The full validation process of a training model includes a clinical validation. Therefore, different imaging modalities were used (sonography and fluoroscopy, fluoroscopy only and iPAD assisted).

The participants had different expertise levels, the most experienced surgeon (n = 1 > 500 PNL), experienced surgeons (n = 2 > 100 PNL), and the residents (n = 5) with different expertise according to their education.

Statistical analyses

All analyses were performed in SAS 9.3 (SAS Institute, Cary, NC). The correlation of puncture time and number of punctures with parameters such as method, and expertise of the surgeon was reviewed. Every surgeon used three different methods, performing a total of 12 experiments; analyses for independent samples were used. Gaussian distribution of the values according to time and the number of trials cannot be expected. We used rank sum tests. To compare two groups, the Mann–Whitney U-Test was used. To compare more than two groups, the Kruskal–Wallis Test was used. Two-sided p < 0.05 were considered statistically significant. If the result of the Kruskal–Wallis Test was significant, the subgroups were compared pairwise using the U-Test.

Results

Table 1 shows the different parameters of the model in comparison to the real-world setting (Face validity).

Table 1.

Face Validity: Comparison of the Model to the Real-World Setting

Compared parameter	Puncture model	Real setting
Punctured structure	Pig kidney/vessel organ complex in anatomically correct position embedded in ballistic gel	Human kidney/flank anatomy
Overlying structures	Ballistic gel	Skin
	Kidney	Subcutaneous tissue
		Thoracolumbar fascia lumbodorsalis muscle/ribs
		Gerota fascia
		Kidney fat
		Kidney
Anatomical structure	Larger number of smaller calices	Smaller number of larger calices
Puncture needle	Real puncture needle used (Cook Medical, Bloomington, IN)	Real puncture needle used (Cook Medical)
Freehand puncture technique	Sonographic Device BK	Sonographic Device BK
	C-Arm, Lithoscope	C-Arm, Lithoscope
	iPAD assisted	iPAD assisted

The technical aspect of the puncture showed almost no differences between the real-world setting and the model. The puncture was done using the same materials and puncture visualization techniques. The main differences are reflected in the anatomical kidney structures and the surrounding tissue. The puncture of a minimally dilatated porcine kidney is more difficult than the puncture of a dilated human kidney, and overlying structures such as ribs and the iliac crest can be cumbersome during the puncture. The main difference of the model lies in the anatomical structures. The perirenal structures are much more easily visible than in other training models,^9

–14 and more closely resemble the real circumstances in anatomy (Fig. 2a, b). The homogeneous structure of the ballistic gel appears not as a disadvantage of the model but rather a perfect situation to visualize the needle tract during the sonographic puncturing process.

FIG. 2.

(a) Ultrasonography of a patient with a hydronephrosis needing a percutaneous drainage of the kidney. Above the kidney, a hyperechoic inhomogenous layer is reflecting the muscular and fatty layers covering the kidney. (b) Ultrasonography-guided puncture using our model. Above the kidney the hypoechoic zone caused by the gel-layer and the puncture needle is displayed as a white line.

To evaluate the content validity of the model, every puncture method used in the real-life scenario was carried out using the model as a surrogate for the human body. The time, number of attempts, and radiation exposure (dose area product) needed to successfully puncture a predefined calix using three different puncture techniques (Ultrasonography, X-ray, and iPAD) was recorded. Figure 3a and b show an X-ray-guided puncture.

FIG. 3.

(a) X-ray-assisted puncture of the kidney in a patient. (b) X-ray-assisted puncture in the model.

To evaluate the construct validity of the model, the differences within each method was measured when the puncture was done by an experienced urologist (three participants) or an inexperienced resident (five participants). Each participant performed 12 punctures using each method, for a total of 96 punctures evaluated.

Puncture time

The time difference to a successful puncture between the sonographic and X-ray-assisted puncture is marginal. The sonography- guided puncture is 31% faster than the iPAD-assisted puncture (hazard ratio = 1.31; p = 0.12). The X-ray-assisted puncture is 36% faster than the iPAD-assisted puncture (hazard ratio = 1.36; p = 0.021). The X-ray-assisted puncture is significantly faster than the iPAD-assisted puncture (Table 2 and Fig. 4).

FIG. 4.

Puncture times of the three different freehand puncture methods in minutes (iPAD-, sonography- and X-ray-assisted puncture). P (unfinished) shows the successful puncture.

Table 2.

Time and Radiation Exposure to a Successful Calyceal Puncture Using Different Methods

Method	Puncture time (minute)—mean (range)	Radiation exposure (mGym²)—mean (range)
Sonography	1.279 (0.2–7.383)	0 (0–91.6)
X-ray	1.225 (0.05–5.533)	8.4 (0–126.7)
iPAD	1.626 (0.2–9.083)	0.8 (0–92)

Some participants used X-ray additional to the sonography-guided puncture, if the puncture was difficult.

Puncture attempts

In terms of number of punctures until being successful, no global difference between the three methods could be shown (Kruskal–Wallis chi-squared = 1.2973, p = 0.5228). Comparing the iPAD with the sonographic method (Wilcoxon W = 4913, p = 0.4165) and iPAD with the X-ray method, there are no statistical differences between these groups (Wilcoxon W = 5024, p = 0.2659).

All participants tried as many times until successful entering the calix. The mean number of attempts of the X-ray-guided puncture was 2 (1–6), using the sonographic method 2 (1–8) and with the iPAD 3 (1–7).

Radiation exposure

The radiation exposure differs significantly between the three puncturing methods (X² = 122.0, p < 0.001). As expected, the X-ray-assisted puncture causes the most exposure to radiation. The iPAD-assisted puncture requires less radiation exposure (Wilcoxon W = 1910, p < 0.001). Sonography-assisted puncture causes significantly lower radiation exposure than the iPAD-assisted puncture (Wilcoxon W = 6427, p < 0.001). Table 2 shows the different radiation exposure according to the three methods.

Puncture time as a function of the expertise level of the surgeon

Comparing the puncture times of all three methods, the most experienced surgeon is 13% faster than the least experienced surgeon (hazard ratio = 1.13, p = 0.42).

The medium experienced surgeons are also 13% (hazard ratio 1.13, p = 0.47) faster than the most experienced surgeon. But there is no significance between the groups. The results are shown in Table 3. The least experienced surgeons needed the most time.

Table 3.

Freehand Puncture-Time in Minutes, for Surgeons with Different Expertise Levels

Expertise	Median (range)
Most experienced surgeon (n = 1)	1.628 (0.2–9.083)
Experienced surgeon (n = 2)	1.279 (0.2–7.383)
Novice (n = 5)	1.434 (0.05–7.0)

Comparing the puncture trials of all methods to the experience level of the surgeons, a clear difference was demonstrated (Kruskal–Wallis chi-squared = 17.392, p < 0.001). The senior surgeon needed the least trials until a successful puncture compared to the residents (Wilcoxon W = 8350; p < 0.001), but also compared to the most experienced surgeon (Wilcoxon W = 1837.5; p < 0.001). Both results are statistically significant.

Comparing the radiation exposure of all three methods the residents used the least radiation dose (Wilcoxon W = 4719, p < 0.001). There was also less exposure to radiation when the senior surgeon did the puncture compared to the most experienced surgeon (Wilcoxon W = 1894 p < 0.001). In summary, there is a statistically significant difference between the three groups.

A good content validity was shown, demonstrating that the model is an effective simulation of freehand guided puncturing of the kidney using different techniques.

Concurrent validity cannot be determined, because up to now there is no gold standard. To determine the predictive validity, a learning curve should be defined. There are no standard learning curves in the literature for comparison.

The educational impact of the system was determined by feedback about the system in clinical workshop settings. The feedback was 100% positive and participants attested to a learning effect, including better puncture performance.

The cost efficiency is showed in detail in Table 4.

Table 4.

Cost Report

Principal materials	Company	Price per sold unit + postage in €	Price per puncture model in €
Kidney/Aorta Model	www.optimist.at	100	100
Ballistic Gel TMP 300 Bloom	www.tmp-products.com	100	20
Ureter catheter	www.imp-medical.de	30	60
IMP-Medical
Food coloring, Crazy Colors, Brauns Heitmann	www.brauns-heitmann.de	5	5
Lock & Lock Box 4.8 L 295 × 230 × 106 mm reusable	www.amazon.de	15	15
Subtotal		250	200
Working hours and energy and ingredients	3 hours	50	150
Total			350

Although the ballistic gel is relatively robust and will not burst under the pressure of the ultrasound probe, the holes caused by the puncture needle will eventually damage the surface of the model.

Since the model is a single use model the durability is limited to a maximum of 20 punctures per model.

Discussion

Since the establishment of the first endourology course at the University of Minnesota in 1982, endourological training models have been needed. Particularly to preserve patient safety, the crucial parts of surgical procedures are practiced on training models. While extracorporeal shockwave lithotripsy has been the dominant treatment for kidney stones for years, there is now a global trend toward percutaneous stone surgery. The main aspect of percutaneous surgery is to establish a skin to kidney tract. The crucial goal is to hit the correct puncture site on the first try. To optimize training in percutaneous puncture techniques, numerous models for the training of percutaneous kidney surgery have been developed.^10

–14 Though some of them are routinely used in training workshops and hands-on training sessions, these models are insufficiently validated and their value in the educational process has not been measured.⁹ It is still not clear which models should be included in effective training protocols and how a curriculum for percutaneous kidney surgery should be designed. A recent study by Noureldin et al. summarized the published training models for percutaneous access and their validation statuses.⁹ Their conclusion was that there is a lack of fully validated training models.

To progress in this field, we validated our newly developed puncture model as completely as possible. The model showed good face, content, and construct validity, positive educational impact, and good durability. Furthermore, we calculated the costs and the time needed to generate the model.

In comparison to other training models, there is the possibility to modify this model to the training needs. New components like three-dimensional plastic kidney models or overlying structures like rips can be embedded easily. All different puncture techniques, such as bulls-eye, triangulation technique, sonographic puncture freehanded or with puncture aid, and many more, can be trained. The model was included sucessfully in several hands-on training courses (Juniorakademie Operationskurs 07/2013 [Resident Training Course], Mannheim, Germany; DGU 09/2013 [German Association of Urology–Annual Meeting] Dresden, Germany; BWK-Kurs-Emergency 05/2014, Leogang Sonography Kurs 01/17) and was also used to validate new puncture techniques, for example, iPAD-assisted puncture.¹⁵

The fast, easy and cheap manufacture of our model is definitely an advantage compared to other models, like the PERC Mentor¹⁶ or the chicken/pig model.¹⁰ The puncture feeling is quite realistic due to the texture of the kidney embedded in ballistic gel. Our model is easy for transportation to training courses, for example. Due to the ballistic gel, our model works best to train ultrasonography-guided punctures.

One specific limitation of our model is the lack to train the puncture under respiration movement, which the PERC Mentor¹⁶ is able to do. The timing of the puncture with a moving kidney can be sometimes crucial. In the end, the anesthetist can stop the respiration for a short period of time to overcome this problem.

The main limitation of the study is the relatively low number of trainees and experts performing the punctures. A Chinese study group evaluated primarily the face validity with a large number of participants.^17
–19 The training model should be integrated in daily practice to define its training value. To develop a training curriculum in PNL, many experts and participants, and their learning curves and the skill transfer in the operating room, are needed. Comparable training devices like the chicken/pig¹⁰ or sandwich-model¹¹ are also not yet fully validated and should be reevaluated according to the different validation parameters. Not only validation parameters, but also side parameters such as ease of production, transportation, feasibility, modifiability, costs, and so on should be evaluated. A good training model should be flexible to use different imaging modalities such as X-ray, ultrasonography, or any other new technology. Although a standard validation process including seven parameters for training and simulation models is already known, the whole process is rarely complete.

Up to now the best validated training tool in percutaneous access is the VR-Simulator PERC Mentor (Simbionix).¹⁶ Though three studies showed a potential benefit of training on this platform, the impact on the operation room transfer remains unclear. This shows the difficulty of observing and evaluation of the learning process. Further studies dealing with this topic are needed.

To optimize the durability of our model, especially for delivery to workshops, a biozide agent can be added to the gel and a Thiel-embalmed kidney/vessel specimen can be used.²⁰ We already tested this combination and found no difference in the material characteristics except a longer durability of the specimen.

Conclusion

The newly developed ballistic gel kidney-puncture model is an ideal training model for a large variety of kidney-puncture techniques. The model demonstrates a good face, content, and construct validity not only for freehand sonographic kidney puncture, but also for X-ray-assisted and iPAD-guided puncture,. It is easy to produce in a reasonable amount of time. All components can be ordered via the internet and some parts are available in a typical urological department. The price to produce the model is moderate and the model is easy to transport.

Footnotes

Author Disclosure Statement

No competing financial interests exist.

Abbreviations Used

References

Türk

, Petřík

, Sarica

, et al. EAU guidelines on diagnosis and conservative management of urolithiasis. Eur Urol, 2016; 69:468–474.

Nagele

, Schilling

, Abastasiadis

, et al. [Minimally invasive percutaneous nephrolitholapaxy (MIP)]. Urologe A, 2008; 47:1066, 1068–1073.

Gao

X-S

, Liao

, CHen

, et al. Different tract sizes of miniaturized percutaneous nephrolithotomy versus retrograde intrarenal surgery: A systematic review and meta-analysis. J Endourol, 2017; 31; 1101–1110.

Michel

, Trojan

, Rassweiler

. Complications in percutaneous nephrolithotomy. Eur Urol, 2007; 51:899–906.

Ritter

, Rassweiler

, Michel

. [Complications in percutaneous lithotomy]. Urologe A, 2014; 53:1011–1016.

de la Rosette

, Laguna

, Rassweiler

, et al. Training in percutaneous nephrolithotomy—A critical review. Eur Urol, 2008; 54:994–1001.

Noureldin

, Fahmy

, Anidjar

, et al. Is there a place for virtual reality simulators in assessment of competency in percutaneous renal access?. World J Urol, 2016; 34:733–739.

Brown

. What is construct validity?. JALT Testing & Evaluatio SIG NL, 2000; 4:8–12.

Noureldin

, Andonian

. Simulation for percutaneous renal access: Where are we?. J Endourol, 2017; 31:S10–SS19.

10.

Häcker

, Wendt-Nordahl

, Honeck

, et al. A biological model to teach percutaneous nephrolithotomy technique with ultrasound- and fluoroscopy-guided access. J Endourol, 2007; 21:545–550.

11.

Imkamp

, von Klot

, Nagele

, et al. New ex-vivo organ model for percutaneous renal surgery. Int Braz J Urol. 2011; 37:388–394.

12.

Jutzi

, Imkamp

, Kuczyk

, et al. New ex vivo organ model for percutaneous renal surgery using a laparoendoscopic training box: The sandwich model. World J Urol, 2014; 32:783–789.

13.

Hunt

, Ristolainen

, Ross

, et al. Low cost anatomically realistic renal biopsy phantoms for interventional radiology trainees. Eur J Radiol, 2013; 82:594–600.

14.

Turney

. A new model with an anatomically accurate human renal collecting system for training in fluoroscopy-guided percutaneous nephrolithotomy access. J Endourol, 2014; 28:360–363.

15.

Muller

, Rassweiler

, Klein

, et al. Mobile augmented reality for computer-assisted percutaneous nephrolithotomy. Int J Comput Assist Radiol Surg, 2013; 8:663–675.

16.

Knudsen

, Matsumoto

, Chew

, et al. A randomized, controlled, prospective study validating the acquisition of percutaneous renal collecting system access skills using a computer based hybrid virtual reality surgical simulator: Phase I. J Urol, 2006; 176:2173–2178.

17.

Zhang

, Yu

, Liu

, et al. Training for percutaneous renal access on a virtual reality simulator. Chin Med J (Engl), 2013; 126:1528–1531.

18.

Zhang

, Ou

, Jia

, et al. Novel biologic model for percutaneous renal surgery learning and training in the laboratory. Urology, 2008; 72:513–516.

19.

Zhang

, Yu

C-F

, Jin

S-H

, et al. Validation of a novel non-biological bench model for the training of percutaneous renal access. Int Braz J Urol. 2014; 40:87–92.

20.

Yiasemidou

, Roberts

, Glassmann

, et al. A multispecialty evaluation of Thiel cadavers for surgical training. World J Surg, 2017; 41:1201–1207.