Effect of measurement procedure errors on assessing lung fluid via remote dielectric sensing system

The study assessed the impact of procedural errors on the remote dielectric sensing system (ReDS), a non-invasive lung fluid assessment technology, in an Asian cohort. Healthy volunteers underwent ReDS measurements following manufacturer’s instructions, with two consecutive measurements one minute apart. A subset of 20 participants had modified procedure settings. Reliability was measured using intraclass correlation coefficient (ICC). The study included 86 healthy volunteers, and all ReDS measurements fell within the recommended normal range. The intra-rater reliability of ReDS measurements was excellent, with an ICC of 0.968. Among the subset of 20 subjects, deviations in height and weight did not significantly affect ReDS values. However, deviations in chest size by ± 3 cm had a noticeable impact on ReDS measures, and incorrect station selection led to fluctuations in ReDS readings. In conclusion, the ReDS system demonstrated excellent intra-rater reliability and applicability in an Asian cohort. Procedural errors, such as chest size measurement and station selection, significantly influenced ReDS measurements. Adherence to standardized operating procedures is crucial to ensure accurate and consistent results. These findings highlight the importance of adherence to manufacturer instructions when utilizing ReDS for lung fluid assessment, thereby enhancing its reliability and clinical applicability.


Study protocol
All ReDS measurements were conducted following manufacturer's instructions.The measurements were performed while the subjects were seated on a long back chair, and a single operator who was certified by the manufacturer carried out the measurements.The study protocol consisted of two parts.In the first part, two consecutive measurements were taken on each participant, with a one-minute interval between them.The readings obtained were then averaged for further analysis.This part of the protocol aimed to assess the performance of the ReDS system in an Asian cohort and determine the intra-rater reliability of the ReDS measures.
In the second part of the study, a subset of 20 subjects was randomly chosen from the cohort.The aim of this part was to assess the impact of procedural errors on the ReDS readings.The subjects underwent serial ReDS measurements with specific modifications to the procedure settings (Fig. 1).In each specified procedure setting, measurements were taken twice with a 1 min interval, and the values were averaged for statistical analysis.Initially, after the standard measurements, ReDS readings were obtained while manipulating the height entry by ± 1 cm, 3 cm, or 5 cm from the true value.Subsequently, ReDS measurements were taken with modifications to the weight entry by ± 1 kg, 3 kg, or 5 kg from the actual value.We then obtained ReDS values after modifying the chest size measurement by ± 1 cm, 3 cm, or 5 cm from the measured value.Finally, ReDS measurements were conducted with the selector set to stations other than the one recommended by the system.

Sample size calculation
The test-retest reliability of ReDS measurement, assessed using the intraclass correlation coefficient (ICC), was expected to be 0.95 10 .The measurement was taken with two repetitions.The minimum acceptable ICC was set

Intra-rater reliability
The average ReDS readings (Table 1) for the first and second measurements in the study cohort were 26% and 27%, respectively.Figure 2 provides a detailed visualization of the changes in ReDS values between the first and second measurements.The ICC between the two measures was 0.968 (95% CI 0.951-0.979),indicating excellent reliability.

Impact of procedure errors on ReDS
In the second part of the study, a total of 20 subjects participated, with an average age of 37 years, including 7 males.The participants had an average body mass index of 25 kg/m 2 .Height deviations within ± 5 cm and weight  www.nature.com/scientificreports/deviations within ± 5 kg did not significantly affect the ReDS readings (Fig. 3A,B, respectively).However, variations of ± 3 cm or ± 5 cm in chest size led to significant alterations in ReDS measurements by approximately ± 2% or ± 4-5%, respectively (Fig. 3C).Decreasing the chest size entry led to an overestimated ReDS value, while increasing the chest size entry resulted in an underestimated ReDS reading.Furthermore, when the standard station was changed to other stations, the ReDS readings deviated from the original values in an unpredictable manner in terms of both direction and magnitude (Fig. 4).

Discussion
In this study, we assessed ReDS measurements in a healthy Asian population, focusing on intra-rater reliability and the impact of procedure errors on ReDS readings.The key findings of our study can be summarized as follows: (a) all healthy Asian volunteers demonstrated ReDS values within the suggested normal range of 20-35%; (b) the intra-rater reliability of ReDS measurements, performed by a single well-trained operator, was found to be excellent; (c) errors in height entry up to ± 5 cm or weight entry up to ± 5 kg did not have a significant impact on ReDS values; (d) deviations in chest size by ± 3 cm or more had a noticeable effect, i.e., approximately ± 2%, on ReDS measures; (e) incorrect station selection also resulted in notable changes in ReDS readings.Overall, our findings suggest that the normal range of the ReDS system is applicable to an Asian cohort, and it exhibits excellent intra-rater reliability.However, manufacturer's instructions should be followed carefully during the ReDS measurements, as procedure errors such as deviation in chest size measurement and error in station selection can significantly impact the accuracy of ReDS readings.
A recent study conducted by Imamura et al. 14 in Japan, along with ours in Taiwan, both involving healthy subjects spanning a wide age range, demonstrated that all ReDS values fell within the manufacturer-recommended normal range.These findings support the validity of ReDS measures in a population with smaller physical stature.Furthermore, Imamura et al. 10 reported excellent intra-rater reliability of the ReDS system, with an ICC ranging from 0.966 to 0.988, which aligned with the ICC of 0.968 observed in our current study.These results suggest that ReDS readings are reproducible when performed by a single operator and indicate the feasibility of using one-time measurements in clinical practice.
In contrast, the study by Imamura et al. 10 also revealed that the inter-rater reliability of ReDS measures, as indicated by the ICC, could be lower (0.683-0.866).This suggests only moderate reliability when different operators are involved.In addition, there was observed variability in median ReDS values of up to 5% between two operators.Notably, the study did not provide a specific rationale or explanation for the significant variation in ReDS measures observed among different operators, highlighting the need for further investigation to better understand and address this variability.
Our study revealed that even minor deviations in the operating procedure can lead to altered ReDS readings, which may contribute to the suboptimal inter-rater reliability observed.Notably, changes in chest size measurements were found to have a significant impact on ReDS values.While the specific computing algorithm employed by the ReDS system was not accessible to us, we observed that increasing or decreasing chest size entry resulted in underestimated or overestimated ReDS readings, respectively.These findings highlight the sensitivity of ReDS measurements to variations in chest size.We recommend recording chest size measures alongside ReDS readings to ensure accuracy and consistency in results.
Additionally, accurate station selection as recommended by the ReDS system is crucial for consistent ReDS measurements.Incorrect placement of the station can result in fluctuations in ReDS values.As ReDS is increasingly utilized in clinical practice 15 , our study reinforces the need to adhere to standardized operating procedures recommended by the manufacturer.This adherence ensures reliable and accurate ReDS measurements, thereby enhancing the accountability and applicability of ReDS in healthcare settings.
A few limitations of this study should be acknowledged and discussed.First, our study was conducted at a single center, which may affect the generalizability of the findings.However, the consistency of our results with previous research on ReDS values in healthy subjects and the intra-rater reliability of the ReDS system partially validates our findings 10,14 .Further studies in multiple centers and diverse populations are needed to confirm the generalizability of our results.Next, the impact of procedure errors on ReDS readings was observed in a population with relatively small stature.Therefore, it remains to be investigated how these errors may affect ReDS measurements in populations with different physical statures.Furthermore, due to study design limitations, we can only infer that height deviations within ± 5 cm and weight deviations within ± 5 kg did not significantly affect the ReDS readings.The impact of more pronounced weight or height errors on the measurements remains uncertain.As one of the pioneering studies in this area, our findings provide valuable insights that can encourage further research by other clinicians to explore this aspect in different populations and study designs.Finally, this study focused exclusively on a healthy cohort, which raises valid concerns about the generalizability of the findings to individuals with comorbidities, such as heart failure.However, we anticipate that the core principle of adhering to standardized operating procedures for ensuring accuracy and consistency in ReDS measurements remains applicable across diverse patient populations.Undoubtedly, further validation studies involving specific patient groups serve as the gold standard for addressing this question.
The ReDS system is increasingly being recognized for its role in heart failure patient management, and our study provides valuable and novel insights into its operation.Our findings highlight the significance of avoiding procedure errors, as they can inadvertently impact the results of ReDS measurements.It is essential to consider a few important factors when conducting these measurements.Accurate chest size measurement is crucial for obtaining reproducible ReDS values, and adherence to the manufacturer's instructions for correct station selection is paramount.By keeping these considerations in mind, the reliability and accuracy of ReDS measurements can be optimized.

Figure 2 .
Figure 2. Changes in remote dielectric sensing (ReDS) values between the first and second measurements, as displayed in the histogram (A) and scatterplot (B).

Figure 3 .
Figure 3. Impact of procedure errors (A) height, (B) weight, and (C) chest size on remote dielectric sensing (ReDS) measurements.Paired t-tests were conducted to compare each measurement with the standard measurement, unless stated otherwise.

Figure 4 .
Figure 4. Effects of station switching on remote dielectric sensing (ReDS) measures.The different colors indicate the direction and magnitude of changes observed in the ReDS readings when the standard station was switched to other stations.