Reducing radiation exposure and costs: CT body scout views with an enhanced protocol versus conventional radiography after shunt surgery

Objective: Ventriculoperitoneal shunt implantation has become standard treatment for cerebrospinal fluid diversion, besides endoscopic third ventriculostomy for certain indications. Postoperative X-ray radiography series of skull, chest and abdomen combined with cranial CT are obtained routinely in many institutions to document the shunt position and valve settings in adult patients. Measures to reduce postoperative radiation exposure are needed, however, there is only limited experience with such efforts. Here, we aim to compare routine postoperative cranial CT plus conventional radiography series (retrospective arm) with cranial CT and body scout views only (prospective arm) concerning both diagnostic quality and radiation exposure. Patients and Methods: After introduction of an enhanced CT imaging protocol, routine skull and abdomen radiography was no longer obtained after VP shunt surgery. The image studies of 25 patients with routine postoperative cranial CT and conventional radiography (retrospective arm of study) were then compared to 25 patients with postoperative cranial CT and CT body scout views (prospective arm of study). Patient demographics such as age, sex and primary diagnosis were collected. The image quality of conventional radiographic images and computed tomography scout views images were independently analyzed by one neurosurgeon and one neuroradiologist. Results: There were no differences in quality assessments according to three different factors determined by two independent investigators for both groups. There was a statistically significant difference, however, between the conventional radiography series group and the CT body scout view imaging group with regard to radiation exposure. The effective dose estimation calculation yielded a difference of 0.05 mSv (two-tailed t-test, p = 0.044) in favor of CT body scout view imaging. Furthermore, the new enhanced protocol resulted in a reduction of cost and the use of human resources.


INTRODUCTION
Ventriculoperitoneal (VP) shunt placement is the most widely used treatment of hydrocephalus diverting the cerebrospinal fluid (CSF) from the cerebral ventricles to the abdominal cavity [1].
Because of a high rate of shunt failure within the first year after shunt implantation, the continuity and position of the shunt system and the pressure setting of programmable valves usually are evaluated postoperatively by imaging studies consisting of X-ray radiographs of the head and neck (anterior-posterior and lateral), chest (anterior-posterior) and abdomen (anteriorposterior and lateral) as well as cranial CT [1,2,3].Some previous studies showed that the use of limited MRI imaging after VP shunt implantation provides adequate visualization of the shunt system while minimizing radiation exposure, however, the access to MRI facilities might be restricted due to accessibility, patient contraindications or cost considerations [4,5].
Furthermore, imaging studies are obtained in patients with suspected shunt dysfunction or shunt infection [6].These secondary X-ray radiographs and cranial CT scans can then be compared to the early postoperative imaging concerning both ventricular width and integrity of the shunt system [7,8].Patients' concerns about ionizing radiation due to radiographic imaging and the fear of malignancies, however, are prevalent [2].With that regard, efforts to reduce postoperative radiation exposure are welcome [9,10], but have been explored only more recently [11].
Previous experimental studies showed that the effective dose as an equivalent of the radiation exposure could be significantly reduced in swine cadaver and in rabbit models when using J o u r n a l P r e -p r o o f whole-body CT views as compared to radiography shunt series for postoperative shunt imaging controls [6,12,13].Diagnostic imaging confidence was considered to be even higher with whole-body CT views in these studies.Furthermore, the image quality of low-dose CT was deemed suitable to visualize the distal catheter position after VP shunt surgery without additional radiation exposure in comparison to added conventional X-ray radiography [14,15].
There is still a lack of studies evaluating the utility of low-dose whole body CT imaging in clinical practice after VP shunt surgery.The present study compares the diagnostic quality of CT body scout views with cranial CT versus conventional X-ray radiography in three categories (overall diagnostic quality, visualization of the shunt system, and contrast of the shunt system versus neighboring tissue) after VP shunt surgery for determination of shunt placement in utilizing a retrospective and a prospective arm.Furthermore, this study compares radiation exposure and procedure-related costs between the two groups.

Study design
The protocol consisted of two arms: 1) retrospective arm with routine postoperative cranial CT and conventional radiography: 25 patients who underwent VP shunt surgery from January, 2021 to October, 2021, and 2) prospective arm with postoperative cranial CT and CT body scout views: 25 patients who underwent VP shunt insertion from October, 2021 to July, 2022.
The CT body scout view algorithm was developed prior to patient recruitment.Inclusion criteria for the present study were primary shunt surgeries and age at surgery 18 years or older.
Exclusion criteria were shunt revision surgeries, ventriculoatrial shunts and patients who were under artificial ventilation.Patient-related characteristics were recorded including age, gender, diagnosis, and imaging characteristics.

Surgical technique
Shunt surgery was performed according to departmental standard techniques as described in detail elsewhere [3,16,17,18,19].Miethke proGAV valves (Potsdam, Germany) consisting of a programmable differential pressure valve and a gravitational unit [3,18] were used in both patient groups.Ventricular catheters were inserted via a frontal precoronary burr hole.In all patients, the distal catheter was implanted in the intraperitoneal cavity via a midline substernal incision.The adjustable valve was set at 6 cm H2O; the gravitational units were nonadjustable (25 cm H2O).To strengthen the indication for shunting, epidural intracranial pressure monitoring was employed preoperatively for two days in patients with idiopathic intracranial hypertension [19].As published previously [16], the AxiEM neuronavigation system was used (Medtronic, Minneapolis, MN, USA) for the implantation of ventricle catheters in these cases.

Postoperative care
On the first postoperative day, cranial CT was performed in all patients to check for hemorrhage and in order to visualize the position of the ventricular catheter.Furthermore, in the retrospective arm of the study "shunt series" (Philips MD Eleva; X-ray) were obtained including anterior-posterior (AP) and lateral radiographs of the head and neck, AP radiographs of the chest, and AP and lateral radiographs of the abdomen (Figure 1).In the prospective arm of the study, cranial CT was supplemented by body scout view imaging of the head, the chest, and abdomen with an additional view for the valve setting (Canon Aquilion One; computer tomograph) (Figure 2).A neuroradiologist reconstructed the depiction of the valve pressure level from the 3D CT data.All imaging data were recorded and managed using PACS (General Electric Centricity Health Care, United States), the routine medical imaging archiving system hosted at Hannover Medical School.

Imaging assessment
The conventional radiography series was obtained with a constant voltage of 120 kV and automatic exposure.The cumulative Dose Area Product (DAP, mGy×cm²) of the five radiographies was converted to the effective dose (mSv) [10].
In the prospective arm of the study, the cumulative Dose Length Product (DLP, mGy×cm) of the CT body scout view images was converted using available conversion factors [19] to the effective dose (mSv).This measure allowed to compare effective doses of the conventional radiography series and the CT body scout view images.The quality of the conventional radiography series and the CT body scout view images was independently evaluated by two observers: one neurosurgeon (ME) and one neuroradiologist (PB) for three different factors: 1) overall diagnostic quality, 2) visualization of the shunt system, and 3) contrast of the shunt system versus neighboring tissue.

Descriptive statistics:
For metric scale data, the mean ± standard deviation is given unless otherwise noted.

Observer reliability:
The data used to analyze agreement consisted of ratings of the qualities of three radiological categories, two radiological diagnostic techniques, and two independent raters.The data enclosed qualities such as overall image assessment, visualization of the shunt system, and J o u r n a l P r e -p r o o f adjacent tissue from CT scout views and radiographic images, thus forming a grid of six data subsets.Data were collected from all 25 patients on a five-point ordinal scale.The computation of agreement in the subsets was conducted by Kendall's coefficient of concordance (Kendall's tau); the interpretation of results referred to Landis and Koch [22].

Patient characteristics
The mean age of patients was 63.4 ± 17.0 years (range: 24-85 years) in the conventional radiography series group and 64.6 ± 15.2 years (range: 20-82 years) in the CT body scout view imaging group.Sixteen patients had idiopathic normal pressure hydrocephalus (iNPH), three had idiopathic intracranial hypertension, and 31 had hydrocephalus malresorptivus (after subarachnoid hemorrhage, tumor resection, or meningitis).The distribution of diagnoses was comparable between the two groups (Table 1).
There were no shunt failures or surgical revisions within one year after surgery.Three patients (6%) developed shunt infections within two years postoperatively.

J o u r n a l P r e -p r o o f
A demographic analysis showed no statistically significant difference between the two groups for age [p: 0.8386], height [p: 0.6617], weight [p: 0.0622], or BMI [p 0.0535]).

Comparative quality of imaging modalities
There was a high interrater concordance.Table 2 shows a summary of the scores of the two evaluators for the different factors.Table 3 shows the analysis of agreement using Kendall's test of accordance (Kendall's tau-statistic) of the two independent raters for the three different factors.For five of six quality factors, substantial high-level agreement was found.Moderate concordance was found for the factor "contrast of the shunt system versus neighboring tissue" for the conventional radiography series.

Costs
The cost of the CT body scout view imaging was less than the cost of the conventional radiography series.According to the clinical management system at Hannover Medical School, the cost of the cranial CT plus the CT body scout view images amounted to 98 €, compared to 145.77 € for the cranial CT plus the conventional radiography series.

DISCUSSION
Our study clearly demonstrates that the CT body scout view imaging protocol to evaluate shunt placement postoperatively resulted in a significant radiation dose reduction as compared to conventional radiography "shunt series".Additionally, our study shows that there is no loss in imaging quality comparing three different measures for routine postoperative control of the shunt system.While theoretically there is a reduction of costs using CT body scout view imaging instead of radiography shunt series, this does not result finally in a different coding according to the German Diagnosis Related Groups (G-DRG) billing system.This study indicates that CT body scout view imaging might actually replace conventional shunt series for postoperative shunt control.This would be advantageous since one of the most critical concerns in the long-term follow-up of patients with VP shunts is the accumulative radiation exposure from several shunt series radiographs and CT scans [6].This would be particularly relevant in the pediatric population where numerous imaging controls can be necessary during J o u r n a l P r e -p r o o f the lifespan of the patient [2,10,23].With that regard it has been estimated that the risk of developing a neoplasm increases 2.4-fold [24,25].A study on 2224 children who had radiation therapy for tinea capitis had a relative risk of 3.6 for the development of basal cell carcinoma during follow-up to 50 years [26].Furthermore, meningioma and other tumors may occur after radiation exposure [27,28,29].
Another option to visualize the shunt system postoperatively is MRI imaging [4,5].Marin et al. [5].showed that employing a rapid MRI-shunt protocol may serve as a viable alternative to CT scans for the majority of pediatric patients with ventricular shunts.However, while this approach significantly reduces radiation exposure, it is essential to consider the trade-off of increased time requirements.Although rapid MRI shunt protocol itself entails only a 7-minute scanning duration, obtaining an MRI necessitates additional time due to stringent screening procedures.Also, the limited access to MRI needs to be considered.
In the conventional radiography series, a dose reduction could result in increased image noise and reduced image quality when reconstructed using traditional filtered back projection (8).
The patients with high BMI normally need higher radiation doses due to the difficulty in evaluating the distal shunt catheter position (15).
Our study confirms previous experimental and clinical pilot studies concerning the equivalence of the imaging quality of CT body scout view imaging of the shunt system [2,30].Further, an orthopedic study by Hey et al. demonstrated that full-body CT scout images are comparable to conventional chest radiographic images for preoperative evaluation of patients undergoing spine surgery showing no significant differences between the two imaging modalities [31].

J o u r n a l P r e -p r o o f
Pala et al. reported that the estimated effective dose was significantly lower with a low-dose CT study protocol as compared to conventional radiography studies in a retrospective series of patients [14].Both, the identification of the abdominal position of the distal shunt catheter and the visualization of the entire shunt system were considered superior with low-dose CT.On the other hand, the identification of the valve settings was judged to be worse with low-dose CT.
Another study used a retrospective chart review of 57 patients who had previous shunt surgery to compare conventional shunt series radiographs with EOS low-dose radiography.The radiation dose was lower with EOS low-dose imaging at a similar field of view than with conventional radiography [2].Furthermore, Ryan et al. [8] demonstrated a significant reduction of radiation dose with a whole-body ultralow-dose CT protocol compared to standard radiographic shunt images in a series of 18 patients after shunt placement.
According to the G-DRG billing system in Germany the use of CT body scout view imaging would not result in true cost savings for shunt control postoperatively.Nevertheless, the net cost of CT body scout view imaging would be less than that of conventional shunt radiography series.Costs differ, however, markedly from country to country.A group from the US indicated that a CT scout scan would be in the range of $236-$366, as compared to $1300-$1547 for conventional radiographic images for shunt control [2].In addition, CT body scout view imaging for shunt control results in an overall reduction of assigning human resources, since there is no need to transport the patient from the CT to the X-ray device in the radiology suite.
In addition, CT body scout view imaging for shunt control results in an overall reduction of assigning human resources, since there is no need to transport the patient from the CT to the Xray device in the radiology suite.This is important with regard to the fact, that in the last few years, a decrease of human resources in many western countries has become a relevant problem.
J o u r n a l P r e -p r o o f applied a before and after approach with a retrospective and a prospective arm.Thus, a future step would be to confirm the results obtained here in a prospective randomized study.Further, because of the relatively low number of patients and the low complication rates, we could not evaluate whether irregularities such as disconnection of the shunt would be reliably detected with CT body scout view imaging.

CONCLUSION
In conclusion, we show that CT body scout view imaging obtained with a standardized imaging protocol provides sufficient imaging quality when compared with conventional shunt series radiography to determine postoperative shunt positioning and valve settings resulting both in lower radiation exposure and a deduction of indirect or direct costs.

Figure 1 :Figure 2 :Figure 3 :
Figure 1: Routine set of postoperative "shunt series" radiographs view imaging as assessed by two independent evaluators.Each factor was evaluated according to a five -point nominal scale: 5 = optimal, 4 = good, 3 = acceptable 2 = partial detection possible, and 1 r n a l P r e -p r o o f

Table 1 :
Patients clinical characteristics in both groups.

Table 2 :
Comparison of the quality of conventional radiography series and CT body scout view imaging as assessed by two independent evaluators.

Table 2 :
Comparison of the quality of conventional radiography series and CT body scout