Sustainability in Dentistry—Insights into Waste Impacts from a Carbon Footprint Comparison Between Conventional and Digital Impression Techniques

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

• The authors should more explicitly articulate the specific novelty of this work compared with existing LCA studies in dentistry. A concise statement highlighting the unique methodological or empirical contribution of Germany-specific primary data, scanner allocation logic, and waste categorization would strengthen the manuscript.
• The choice of the “positive dental model” physical stone cast as the functional unit versus the digital CAD model is reasonable, but its clinical and practical equivalence should be justified more explicitly. The authors should clarify whether differences in downstream fabrication routes, casting vs. CAD/CAM, could affect functional comparability.
• The exclusion of patient and staff travel, building energy consumption, and shared instrument production should be more rigorously justified. A brief qualitative discussion on how the inclusion of these factors might influence relative results would improve transparency and avoid misinterpretation.
• The usage-based allocation approach for intraoral scanner manufacturing is innovative but complex. The rationale for selecting operating time and impression frequency as allocation parameters should be explained more clearly and compared briefly with alternative allocation approaches, such as economic or functional allocation.
• Key assumptions regarding scanner lifespan 880 days and daily utilization rates strongly affect the ecological break-even analysis. These assumptions should either be supported by empirical evidence or presented as scenario-based ranges to enhance robustness.
• The suitability of Ecoinvent datasets for hazardous versus municipal waste incineration is questioned later in the manuscript, but this issue should be addressed earlier in the Methods section, as it has a substantial influence on waste-related emissions and on the interpretation of results.
• Although Monte Carlo simulation is used to quantify uncertainty in PCF results, uncertainty arising from methodological choices, allocation rules, waste classification, and database selection is not addressed. A qualitative uncertainty discussion aligned with ISO 14044 guidance would strengthen the study.
• The exclusive focus on Global Warming Potential should be more critically acknowledged as a limitation. Other impact categories frequently cited in the literature review, such as human toxicity, ecotoxicity, and water use, are omitted. The authors should justify this choice and briefly discuss potential trade-offs.
• Primary data collection is based on measurements from 15 patients within a single exemplary practice. The authors should discuss representativeness and potential variability across different practice types and clinical settings, particularly with respect to generalizability.
• Cleaning and disinfection processes are major contributors to environmental impact in both analogue and digital workflows. This important finding deserves deeper discussion, including practical mitigation strategies such as alternative disinfectants, reusable cleaning materials, or optimized cleaning protocols.
• The ecological break-even analysis is a strong component of the manuscript but may be difficult for practitioners to interpret. The authors should simplify their presentation and translate results into clearer, practice-oriented guidance.
• Although a conflict of interest statement is provided, the role of the industry partner in data provision and device selection should be described more transparently to avoid perceived bias and enhance credibility.
• The graphical presentation of the results of Figures 3 and 4 is information-rich but visually dense. Improvements in readability, such as simplified color schemes, larger fonts, clearer labels, or the addition of a summary comparison table, would enhance accessibility.
• The discussion would benefit from a stronger linkage to policy and regulatory implications, particularly in the context of German and EU circular economy strategies. Explicitly outlining how the findings could inform procurement, waste regulation, or clinical guidelines would increase impact.
• The limitations and future research directions should be more explicitly structured, including constraints such as country specificity, reliance on a single scanner model, and exclusion of downstream fabrication stages, as well as recommendations for multi-impact, multi-country, and clinically integrated future LCAs.

Author Response

Response to Reviewer Comments

 

Comment 1: The authors should more explicitly articulate the specific novelty of this work compared with existing LCA studies in dentistry. A concise statement highlighting the unique methodological or empirical contribution of Germany-specific primary data, scanner allocation logic, and waste categorization would strengthen the manuscript.

• Response 1: The final paragraph of the introduction has been improved to point out the novel contributions of this study.

 

Comment 2: The choice of the “positive dental model” physical stone cast as the functional unit versus the digital CAD model is reasonable, but its clinical and practical equivalence should be justified more explicitly. The authors should clarify whether differences in downstream fabrication routes, casting vs. CAD/CAM, could affect functional comparability.

• Response 2: This issue has been clarified in the final paragraph of the "Goal and Scope Definition" chapter.

 

Comment 3: The exclusion of patient and staff travel, building energy consumption, and shared instrument production should be more rigorously justified. A brief qualitative discussion on how the inclusion of these factors might influence relative results would improve transparency and avoid misinterpretation.

• Response 3: A brief qualitative discussion on how the inclusion of travel and building energy consumption affects the results has been added to the "Goal and Scope" chapter.

 

Comment 4: The usage-based allocation approach for intraoral scanner manufacturing is innovative but complex. The rationale for selecting operating time and impression frequency as allocation parameters should be explained more clearly and compared briefly with alternative allocation approaches, such as economic or functional allocation.

• Response 4: We have added a further calculation step and a parameter table to clarify the calculation. Furthermore, we added the following justification: "This approach was chosen because economic allocation, based on cost shares, does not adequately reflect the physical wear and usage of the device. Similarly, functional allocation, which distributes impacts equally across all applications, would ignore differences in duration and frequency of use, which vary among dentists and dental practices focused on specific procedures, such as impressions."

 

Comment 5: Key assumptions regarding scanner lifespan 880 days and daily utilization rates strongly affect the ecological break-even analysis. These assumptions should either be supported by empirical evidence or presented as scenario-based ranges to enhance robustness.

• Response 5: Predicting the exact lifespan is challenging as the technology is relatively new. However, scenarios and explanatory sentences have been added following the introduction of the parameters to enhance the robustness of the model.

 

Comment 6: The suitability of Ecoinvent datasets for hazardous versus municipal waste incineration is questioned later in the manuscript, but this issue should be addressed earlier in the Methods section, as it has a substantial influence on waste-related emissions and on the interpretation of results.

• Response 6: The following has been added to the "Goal and Scope" chapter: Waste was categorized into cardboard, glass waste, wastewater, municipal waste, and hazardous waste, with the classification based on the German Waste Catalogue (AVV). Disinfectants and contaminated waste were classified as hazardous under code 180106*. The two Ecoinvent datasets ‘municipal solid waste (Germany), market for municipal solid waste, incineration’ and ‘hazardous waste, for incineration (Europe without Switzerland, market for)’ were used for this purpose, with the latter having a five times higher PCF than the municipal waste dataset. In this study, this classification applies only to materials in direct contact with the patient, specifically the soaked cloths and paper towels used to clean the scanner tips (digital) and impression trays (analog), as well as the disinfection solutions used for alginate and silicone models (analog). All other materials were disposed of as municipal solid waste, while their corresponding packaging materials were managed according to German recycling regulations.

 

Comment 7: Although Monte Carlo simulation is used to quantify uncertainty in PCF results, uncertainty arising from methodological choices, allocation rules, waste classification, and database selection is not addressed. A qualitative uncertainty discussion aligned with ISO 14044 guidance would strengthen the study.

• Response 7: We have added a paragraph addressing the uncertainty analysis at the end of the "Results" chapter.

 

Comment 8: The exclusive focus on Global Warming Potential should be more critically acknowledged as a limitation. Other impact categories frequently cited in the literature review, such as human toxicity, ecotoxicity, and water use, are omitted. The authors should justify this choice and briefly discuss potential trade-offs.

• Response 8: We have added a paragraph at the end of the article to address these limitations and discuss potential trade-offs.

 

Comment 9: Primary data collection is based on measurements from 15 patients within a single exemplary practice. The authors should discuss representativeness and potential variability across different practice types and clinical settings, particularly with respect to generalizability.

• Response 9: We have added a paragraph to the "Discussion" chapter to address the representativeness and variability of the data.

 

Comment 10: Cleaning and disinfection processes are major contributors to environmental impact in both analogue and digital workflows. This important finding deserves deeper discussion, including practical mitigation strategies such as alternative disinfectants, reusable cleaning materials, or optimized cleaning protocols.

• Response 10: This has now been addressed towards the end of the "Discussion" chapter.

 

Comment 11: The ecological break-even analysis is a strong component of the manuscript but may be difficult for practitioners to interpret. The authors should simplify their presentation and translate results into clearer, practice-oriented guidance.

• Response 11: We have added a few sentences to translate the findings into practical, practice-oriented guidance.

 

Comment 12: Although a conflict of interest statement is provided, the role of the industry partner in data provision and device selection should be described more transparently to avoid perceived bias and enhance credibility.

• Response 12: We have slightly expanded our disclosure for clarity. While the core research was conducted independently, the industry partner provided detailed information about the formula and therefore weight shares of the individual materials used in the impression taking materials.

 

Comment 13: The graphical presentation of the results of Figures 3 and 4 is information-rich but visually dense. Improvements in readability, such as simplified color schemes, larger fonts, clearer labels, or the addition of a summary comparison table, would enhance accessibility.

• Response 13: We have reworked Figures 3 and 4 for better clarity and added a summary comparison table to the "Results" chapter to enhance accessibility.

 

Comment 14: The discussion would benefit from a stronger linkage to policy and regulatory implications, particularly in the context of German and EU circular economy strategies. Explicitly outlining how the findings could inform procurement, waste regulation, or clinical guidelines would increase impact.

• Response 14: We have followed the reviewer’s suggestion and integrated a new section on policy implications, including a corresponding paragraph in the "Discussion" chapter.

 

Comment 15: The limitations and future research directions should be more explicitly structured, including constraints such as country specificity, reliance on a single scanner model, and exclusion of downstream fabrication stages, as well as recommendations for multi-impact, multi-country, and clinically integrated future LCAs.

• Response 15: We would like to thank the reviewer for this insightful suggestion. Following the extensive revisions made in response to the previous comments, we have carefully integrated all mentioned constraints—including country specificity and the single scanner model—directly into the relevant sections of the discussion. Regarding the downstream fabrication stages, we did actually include processes such as CNC milling, casting, or injection molding; however, these were based on the total amount of material provided by the recycling pass document for the Primescan (Source [35]) instead of individual parts. We now describe this more explicitly in Chapter 3.2.2 as well.

Reviewer 2 Report

Comments and Suggestions for Authors

Dear authors, this manuscript examines a critical sustainability concern in dentistry by comparing the product carbon footprint (PCF) of analogue versus digital workflows for a single-tooth crown, utilizing a life cycle assessment (LCA). The study is supported by primary gravimetric measurements collected in-house and detailed modeling of actual clinical procedures, including cleaning and waste disposal. The primary limitations involve issues with clarity and internal consistency of certain parameters, notably (i) the allocation of scanner manufacturing impacts based on usage, and (ii) some energy input data. While these limitations do not undermine the overall validity of the research, they do impact the reproducibility and robustness of the main comparison (analogue approximately 1620 g CO₂-eq versus digital approximately 550 g CO₂-eq). 

 

The title is suitable and precise. If the main outcome relates to comparing PCF and GWP, it’s recommended to use consistent terminology throughout the document, such as choosing "PCF" or the more general term "environmental impacts," especially if only GWP is highlighted in the Results and Discussion sections.

 

The Abstract effectively outlines the objective, scope (“cradle-to-grave”), and main GWP results. It would be beneficial to add a sentence explicitly stating key exclusions or assumptions within the system boundary, such as the exclusion of travel and facility energy, to prevent misunderstandings when readers compare your data with other studies. If a Monte Carlo uncertainty analysis has been performed, consider including a confidence interval or at least mentioning that uncertainty ranges have been assessed. 

The Introduction highlights the importance of Life Cycle Assessment (LCA) and stresses the need for quantitative research in dental workflows. It is recommended to improve the concluding paragraph by clearly stating the novel contributions of this study, including the main in-practice measurements, classification of German waste, and the device amortization approach. While the use of Generative AI for language polishing is openly disclosed, this should be briefly mentioned and aligned with the journal's language policies. 

Materials and Methods constitute the most critical section to enhance for reproducibility. Concerning the goal, scope, functional unit, and comparability: you define the functional unit as the “positive model” (stone cast versus virtual CAD model). This choice is justified, particularly for workflows that depend on CAM; however, I recommend a brief clarification: what proportion of your intended clinical pathways actually produce these truly functionally equivalent outputs (e.g., analogue workflows that may still be digitized later)?

Secondly, regarding the System boundary and exclusions: the boundary definition is clear (patient seated; travel/facility energy excluded; single-use components cradle-to-grave; reusable instrument production excluded while cleaning is included). Please consider adding a brief “Implications” note: readers should not compare your absolute PCFs to studies that include travel and building energy without proper adjustment. 

Third, waste categorization—an essential factor—entails using the German Waste Catalogue (AVV) to classify waste, including identifying contaminated waste and disinfectants as hazardous substances (e.g., 180106*). This classification significantly influences outcomes; therefore, kindly include a succinct table that correlates each major waste stream (such as gloves, wipes or paper towels, dental stone, impression material residues) with the respective disposal route specified in the modelling process. 

 Fourth, there is a significant clarity issue regarding the allocation of scanner manufacturing impacts. The paper presents two values that could be easily confused: an allocation share of 0.014% (based on lifespan and usage assumptions) and a “proportion of device usage attributable to the single-crown impression” of 6%, which is used to calculate per-impression emissions and break-even points. These may refer to different concepts, such as per-impression share versus device “on-time" share, but this distinction is not clearly stated. I suggest adding a parameter table that clearly defines each symbol (α, e, f, t, tₒₙ, T), including units, values, and brief explanations. Also, showing a simple, step-by-step calculation from the manufacturing impact to the per-impression value would enhance understanding. 

Fifth, it's important to consider energy inputs, which may relate to unit or scale issues. Some recorded energy values—like "Pentamix 3 … consumes 1.5 kWh per application”—seem unusually high for short mixing events. This could be due to factors such as (i) standby energy inclusion, (ii) longer measurement periods, or (iii) a unit mismatch between Wh and kWh. Please specify your measurement method, including the type of meter, measurement intervals, and if standby energy was included. Also, include a plausibility check, such as multiplying runtime by rated power. 

Sixth, with regard to the databases and the impact assessment methodology, the utilization of ecoinvent and ReCiPe 2016 (H) is suitable; however, if the manuscript predominantly reports GWP/PCF, it is advisable to either (a) explicitly focus the interpretation on GWP or (b) include a supplementary table featuring 2–4 additional categories to substantiate the choice of a broader LCIA method. 

Seventh, the uncertainty analysis (Monte Carlo) entails running 20,000 simulations using distributions, with the mean and confidence intervals analyzed. These uncertainty results are currently not displayed in the main results section. Please add the mean ± 95% CI for the main scenarios (analogue, digital average, digital tip variants) and highlight the top 3–5 factors contributing to variability. 

The Results section’s strengths include clear headline values and identifying major contributors such as waste avoidance in digital and device manufacturing share. I recommend adding a brief reconciliation table to demonstrate how the largest process contributions sum up to the reported totals, helping readers validate the results and reducing the 'black box” perception of the model. While the break-even table is helpful, it relies heavily on allocation assumptions. After clarifying α/e/f/t/T as suggested above, consider including a short, 2–3 line explanation on how to read Table 1 and explicitly state which assumptions are being held constant. 

The discussion highlights waste avoidance as a key benefit of digital workflows and points out the ecological amortisation dilemma related to equipment-heavy solutions. Since you explicitly question the validity of the hazardous waste dataset—which has a five times higher PCF than municipal waste despite similar collection and treatment—conducting a scenario analysis would greatly improve credibility: Scenario A represents the current method as modeled; Scenario B considers contaminated waste treated like municipal waste in a proxy incineration; and Scenario C involves a mixed approach based on material type with proper justification. This analysis will show if the roughly threefold difference remains under realistic waste modeling options. 

The Conclusion is generally supported by your model outputs, but its robustness depends on clarifying the allocation and energy issues discussed earlier. It might be useful to explicitly state that the results rely on (i) waste classification assumptions and (ii) scanner utilisation levels. 

With regard to data availability, transparency, and conflicts of interest, the manuscript indicates reliance on manufacturer inputs “wherever possible” and identifies Dentsply Sirona as a project partner and primary data supplier. For enhanced transparency, a brief statement should be included to describe how modelling decisions were maintained independently, such as through author-controlled parameter selection, sensitivity analysis, or acknowledgment of limitations. If sharing raw data is not feasible, consider providing an anonymised, aggregated Life Cycle Inventory (LCI) table (detailing mass and energy totals per impression) along with a list of datasets utilized. This approach should aim to improve reproducibility while respecting any applicable constraints. 

This study investigates an important question about sustainability in dental procedures and provides a clear, practical comparison between analogue and digital impression methods. To strengthen the credibility of the results, it would be helpful to clarify the assumptions about scanner allocation and verify key energy use and waste management factors, ideally with a short sensitivity or scenario analysis. Making these adjustments would improve the transparency, reproducibility, and publication readiness of the manuscript.

Author Response

Response to Reviewer 2 Comments

Comment 1: The Abstract effectively outlines the objective, scope (“cradle-to-grave”), and main GWP results. It would be beneficial to add a sentence explicitly stating key exclusions or assumptions within the system boundary, such as the exclusion of travel and facility energy, to prevent misunderstandings when readers compare your data with other studies. If a Monte Carlo uncertainty analysis has been performed, consider including a confidence interval or at least mentioning that uncertainty ranges have been assessed.

• Response 1: Key exclusions regarding travel and facility energy have been added to the abstract. Furthermore, we have included a mention of the uncertainty analysis.

Comment 2: The Introduction highlights the importance of Life Cycle Assessment (LCA) and stresses the need for quantitative research in dental workflows. It is recommended to improve the concluding paragraph by clearly stating the novel contributions of this study, including the main in-practice measurements, classification of German waste, and the device amortization approach. While the use of Generative AI for language polishing is openly disclosed, this should be briefly mentioned and aligned with the journal's language policies.

• Response 2: The final paragraph of the introduction has been revised to explicitly state the novel contributions of this study.

Comment 3: Materials and Methods constitute the most critical section to enhance for reproducibility. Concerning the goal, scope, functional unit, and comparability: you define the functional unit as the “positive model” (stone cast versus virtual CAD model). This choice is justified, particularly for workflows that depend on CAM; however, I recommend a brief clarification: what proportion of your intended clinical pathways actually produce these truly functionally equivalent outputs (e.g., analogue workflows that may still be digitized later)?

• Response 3: This issue has been clarified in the final paragraph of the "Goal and Scope Definition" chapter.

Comment 4: Secondly, regarding the System boundary and exclusions: the boundary definition is clear... Please consider adding a brief “Implications” note: readers should not compare your absolute PCFs to studies that include travel and building energy without proper adjustment.

• Response 4: A brief qualitative discussion on how the inclusion of travel and building energy consumption might influence the results has been added to the "Goal and Scope" chapter. Additionally, a corresponding note has been included in the "Results" chapter.

Comment 5: Third, waste categorization—an essential factor—entails using the German Waste Catalogue (AVV) to classify waste... kindly include a succinct table that correlates each major waste stream with the respective disposal route specified in the modelling process.

• Response 5: We have clarified the waste classification in the "Goal and Scope" chapter. The hazardous classification applies specifically to materials in direct patient contact (e.g., soaked cloths/paper towels for cleaning scanner tips and impression trays) and disinfection solutions. All other materials are categorized as municipal waste or managed via German recycling regulations. Furthermore, specific waste types are defined for each process within Chapter 2.2.1.

Comment 6: Fourth, there is a significant clarity issue regarding the allocation of scanner manufacturing impacts... I suggest adding a parameter table that clearly defines each symbol (α, e, f, t, ton, T), including units, values, and brief explanations. Also, showing a simple, step-by-step calculation from the manufacturing impact to the per-impression value would enhance understanding.

• Response 6: A parameter table and an additional step-by-step calculation have been included. We have also added a brief discussion regarding alternative allocation methods.

Comment 7: Fifth, it's important to consider energy inputs... Some recorded energy values—like "Pentamix 3 … consumes 1.5 kWh per application”—seem unusually high for short mixing events. Please specify your measurement method, including the type of meter, measurement intervals, and if standby energy was included.

• Response 7: Thank you for identifying this error. The value should have been 1.5 Wh instead of 1.5 kWh; this has been corrected in the text. The underlying calculations were performed correctly using the proper units. Measurements were taken using a Voltcraft SEM5000, specifically for the duration of the mixing process without including standby energy.

Comment 8: Sixth, with regard to the databases and the impact assessment methodology... it is advisable to either (a) explicitly focus the interpretation on GWP or (b) include a supplementary table featuring 2–4 additional categories.

• Response 8: We have clarified the exclusive focus on GWP within the "Methods" chapter.

Comment 9: Seventh, the uncertainty analysis (Monte Carlo) entails running 20,000 simulations using distributions... Please add the mean ± 95% CI for the main scenarios... and highlight the top 3–5 factors contributing to variability.

• Response 9: Because the chosen distribution functions are asymmetrical, the Monte Carlo simulations produce mean values that differ slightly from the static results published in this paper. To avoid confusing the reader with varying figures without a disproportionately extensive methodological explanation, we have opted not to change the main results. However, we have added a dedicated paragraph at the end of the "Results" chapter to address this uncertainty and the factors contributing to variability.

Comment 10: I recommend adding a brief reconciliation table to demonstrate how the largest process contributions sum up to the reported totals... consider including a short, 2–3 line explanation on how to read Table 1 and explicitly state which assumptions are being held constant.

• Response 10: We have followed the reviewer’s suggestion to improve accessibility by adding a detailed explanation for Table 1. This new paragraph explicitly describes how to interpret the relationship between the utilization share and the total volume of impressions, while clarifying which parameters are held constant for the amortization model.

Comment 11: The discussion highlights waste avoidance as a key benefit of digital workflows and points out the ecological amortisation dilemma related to equipment-heavy solutions. Since you explicitly question the validity of the hazardous waste dataset—which has a five times higher PCF than municipal waste despite similar collection and treatment—conducting a scenario analysis would greatly improve credibility: Scenario A represents the current method as modeled; Scenario B considers contaminated waste treated like municipal waste in a proxy incineration; and Scenario C involves a mixed approach based on material type with proper justification. This analysis will show if the roughly threefold difference remains under realistic waste modeling options. 

• Response 11: We thank the reviewer for this suggestion. However, we have opted not to include a quantitative scenario analysis for the different waste treatment routes for the following reason: Standard LCA datasets for waste incineration (such as those in Ecoinvent) are modeled based on average municipal or hazardous waste mixes, reflecting specific average material compositions and carbon contents. These average values dictate the resulting emissions during incineration. Since specific datasets for the chemical composition and precise carbon release of dental-specific materials (e.g., specific elastomers or gypsum) are currently unavailable in established databases, a multi-scenario analysis would rely on the same underlying average emission factors. Instead of introducing potential "pseudo-precision" through such scenarios, we have expanded our qualitative discussion.

Comment 12: The Conclusion is generally supported by your model outputs... It might be useful to explicitly state that the results rely on (i) waste classification assumptions and (ii) scanner utilisation levels.

• Response 12: The discussion has been updated to explicitly state that the results are sensitive to waste classification assumptions and scanner utilization levels.

Comment 13: For enhanced transparency, a brief statement should be included to describe how modelling decisions were maintained independently... consider providing an anonymised, aggregated Life Cycle Inventory (LCI) table.

• Response 13: We appreciate the reviewer's focus on transparency. Regarding the independence of the modeling, we believe that with the extensive revisions made to this manuscript, the transparent disclosure of our industry partnership and the addition of a qualitative sensitivity discussion the academic integrity and independence of our study are now clearly established. We feel that the current depth of methodological detail provides a robust basis for the conclusions drawn.

Regarding the provision of an aggregated Life Cycle Inventory (LCI) table, we unfortunately cannot provide further detailed inventory data. Our industry partners, who provided essential technical specifications and material formulations, do not permit the publication of even aggregated LCI data due to proprietary constraints and confidentiality agreements.

Comment 14: To strengthen the credibility of the results, it would be helpful to clarify the assumptions about scanner allocation and verify key energy use and waste management factors, ideally with a short sensitivity or scenario analysis.

• Response 14: In this revised version, we have now strengthened the transparency and reproducibility of our results. Specifically, we have included a more detailed explanation of the scanner allocation logic and clarified the underlying parameters. Furthermore, we have incorporated a brief sensitivity analysis.

Round 2

Reviewer 1 Report

Comments and Suggestions for Authors

Accepted as authors revised well and now scientifically sound 

Author Response

Thanks, but it says "I would not like to sign my review report"

Reviewer 2 Report

Comments and Suggestions for Authors

Dear author, the revised manuscript offers a clearer and more practically relevant comparison of LCA/PCF between analogue and digital impression workflows for a single-tooth crown case. It includes a well-defined functional unit (positive model) and transparent boundary choices. The update improves interpretability by expanding the Discussion to cover waste regulation context, ecological break-even points, and generalisability limitations, while also presenting uncertainty results more clearly. The main conclusion—digital ≈ 550 g CO₂-eq versus analogue ≈ 1620 g CO₂-eq—remains plausible based on current assumptions, though its robustness depends on waste classification/disposal modeling (especially hazardous waste incineration factors) and the scanner manufacturing allocation and usage scenario. These points are now acknowledged in the revised limitations, marking a meaningful improvement.   

However, some points need clarification:

First, you now describe the allocation through equations and link scanner burden to utilisation and break-even points (Table 1), which is helpful. However, the manuscript still risks confusing the usage share for single-crown impressions (~6%) with the manufacturing allocation fraction per impression (~0.014% / ~0.01377%). Please add a brief parameter/notation table (αᵢ, fᵢ, tᵢ, t_d, T, E_t, eᵢ; with units and explanations). Also, include a short worked example in the Supplementary Materials that step-by-step calculates the 150.8 g CO₂-eq per impression figure. 

Secondly, the revised Discussion appropriately emphasizes that the hazardous waste dataset possesses a Product Carbon Footprint (PCF) approximately five times higher than that of municipal waste. It also queries whether this magnitude is indeed realistic, considering the similar collection and treatment processes involved. Given the significant influence of waste incineration on the analogue footprint, I recommend incorporating a straightforward sensitivity or scenario analysis—such as a descriptive paragraph accompanied by a figure or table. The Base case should adhere to the current methodology (hazardous waste including contaminated wipes, towels, disinfectants, etc.); an Alternative case could simulate contaminated waste using a municipal incineration proxy or a combined factor, justified as a "bounding" scenario. This approach will illustrate whether the conclusion that the PCF is approximately "~3× lower” remains consistent under plausible variations in waste factors. Such an interpretation aligns well with the overall analysis. To further augment the utility for readers, please include the numeric values corresponding to the 10th–90th percentiles (or 95% confidence intervals) for each primary scenario within the Results section, rather than solely in the supplementary materials graphic. 

Third, the Abstract is clear and well-balanced, effectively highlighting the significance of high utilisation for ecological gains. Please consider including an explicit sentence mentioning that the results are specific to the country context (German energy mix and waste standards) to avoid overgeneralisation. Although this point is already mentioned later in the limitations, reinforcing it in the Abstract would ensure consistency.  

Fourth in Introduction The novelty claim is stronger now (in-practice measurements + German waste classification + amortisation approach).   Minor style: tighten a few long background paragraphs so the specific gap and your contribution emerge earlier.

Fifth, the Materials and Methods section, including the goal/scope and the exclusion of travel/facility energy, is well-justified and suitable for isolating procedural differences. The methods are now more reproducible, and the allocation equations are beneficial. Please clarify units consistently throughout (e.g., Wh vs. kWh for device energy per application). The revised text appears more credible than previous versions; maintain this clarity consistently across all device-energy statements. 

Sixth, The Results are clear: digital methods reduce PCF; wipe-disinfectable tips are the most effective; and single-use and autoclave options show similar performance. Table 2 (Top 10 contributors) adds transparency and backs your interpretation, including details on gloves, hazardous chemicals, and waste outputs. Please complete the missing percentage sentence (“accounting for approximately percent…”) to make the statement complete and professional. 

Seventh, the updated Discussion is significantly clearer: it explains burden shifting from materials to equipment, emphasises the difference in 'error/rework' penalties between workflows, and frames the break-even point as a practical decision-making tool. Limitations are now clearly stated and appropriate: they include generalisability from a single site with 15 patients, omitted impact categories, reliance on hazardous waste factors and utilisation assumptions, and country-specific factors. I recommended that if the policy-oriented paragraph (EU Circular Economy Action Plan / KrWG) is kept, it should be concise and directly linked to your quantitative results and waste-prevention strategies.  

Finally, the conclusions align with the results and limitations; it is good that you explicitly acknowledge potential trade-offs in toxicity/resource depletion for electronics and recommend multi-criteria LCA as future work.  Overall, the revised manuscript is nearing completion for publication; however, it would benefit from (i) a clearer and more audit-friendly presentation of the allocation parameters, (ii) a concise sensitivity analysis regarding hazardous-waste modelling, and (iii) the inclusion of explicit numeric uncertainty intervals within the Results section.    

Author Response

Response to Author 2, Round 2

Comment: First, you now describe the allocation through equations and link scanner burden to utilisation and break-even points (Table 1), which is helpful. However, the manuscript still risks confusing the usage share for single-crown impressions (~6%) with the manufacturing allocation fraction per impression (~0.014% / ~0.01377%). Please add a brief parameter/notation table (αᵢ, fᵢ, tᵢ, t_d, T, E_t, eᵢ; with units and explanations). Also, include a short worked example in the Supplementary Materials that step-by-step calculates the 150.8 g CO₂-eq per impression figure. 

• Response: Sorry, now I undertand the confusion. For clarification I changed the sentence giving the final result of 150.8 so that it should be more clear. Both a calculation step and a notation table have reworked as well. Thank you for the suggestion regarding the Supplementary Materials. After review, we have decided to keep the calculation details and the notation table directly within the main manuscript instead.

 

Comment: Secondly, the revised Discussion appropriately emphasizes that the hazardous waste dataset possesses a Product Carbon Footprint (PCF) approximately five times higher than that of municipal waste. It also queries whether this magnitude is indeed realistic, considering the similar collection and treatment processes involved. Given the significant influence of waste incineration on the analogue footprint, I recommend incorporating a straightforward sensitivity or scenario analysis—such as a descriptive paragraph accompanied by a figure or table. The Base case should adhere to the current methodology (hazardous waste including contaminated wipes, towels, disinfectants, etc.); an Alternative case could simulate contaminated waste using a municipal incineration proxy or a combined factor, justified as a "bounding" scenario. This approach will illustrate whether the conclusion that the PCF is approximately "~3× lower” remains consistent under plausible variations in waste factors. Such an interpretation aligns well with the overall analysis. To further augment the utility for readers, please include the numeric values corresponding to the 10th–90th percentiles (or 95% confidence intervals) for each primary scenario within the Results section, rather than solely in the supplementary materials graphic. 

• Response: Good idea. I added a table showing the impact of alternative waste scenarios in the discussion section. Please note that due to current administrative changes in university IT-licensing, we do not have immediate access to the simulation software for generating new stochastic models within the revision window. On top of that it might be confusing anyway since the mean value of the probalistic Monte Carlo model is different to the deterministic model (as explained below figure 5) which is in focus in this paper.

 

Comment: Third, the Abstract is clear and well-balanced, effectively highlighting the significance of high utilisation for ecological gains. Please consider including an explicit sentence mentioning that the results are specific to the country context (German energy mix and waste standards) to avoid overgeneralisation. Although this point is already mentioned later in the limitations, reinforcing it in the Abstract would ensure consistency.  

• Response: Added as proposed.

 

Comment: Fourth in Introduction The novelty claim is stronger now (in-practice measurements + German waste classification + amortisation approach).   Minor style: tighten a few long background paragraphs so the specific gap and your contribution emerge earlier.

• Response: Reworked a few introduction paragraphs as proposed

 

Comment: Fifth, the Materials and Methods section, including the goal/scope and the exclusion of travel/facility energy, is well-justified and suitable for isolating procedural differences. The methods are now more reproducible, and the allocation equations are beneficial. Please clarify units consistently throughout (e.g., Wh vs. kWh for device energy per application). The revised text appears more credible than previous versions; maintain this clarity consistently across all device-energy statements. 

• Response: We thank the reviewer for the positive feedback regarding the reproducibility and the allocation equations. We have carefully re-checked the entire manuscript for unit consistency. We have decided to use Watt-hours (Wh) consistently for all device-energy statements per application. This choice was made to avoid several leading zeros (which would occur using kWh) and thus to ensure better readability and clarity for the reader at the procedural level. We have verified that no 'kWh' or inconsistent energy units remain in the text

 

Comment: Sixth, The Results are clear: digital methods reduce PCF; wipe-disinfectable tips are the most effective; and single-use and autoclave options show similar performance. Table 2 (Top 10 contributors) adds transparency and backs your interpretation, including details on gloves, hazardous chemicals, and waste outputs. Please complete the missing percentage sentence (“accounting for approximately percent…”) to make the statement complete and professional. 

• Response: We have addressed the oversight regarding the missing percentage value. The sentence in the Results section (Section 3.2) has been completed, now explicitly stating that the primary contributors account for approximately 20.2%.

 

Comment: Seventh, the updated Discussion is significantly clearer: it explains burden shifting from materials to equipment, emphasises the difference in 'error/rework' penalties between workflows, and frames the break-even point as a practical decision-making tool. Limitations are now clearly stated and appropriate: they include generalisability from a single site with 15 patients, omitted impact categories, reliance on hazardous waste factors and utilisation assumptions, and country-specific factors. I recommended that if the policy-oriented paragraph (EU Circular Economy Action Plan / KrWG) is kept, it should be concise and directly linked to your quantitative results and waste-prevention strategies.  

• Response: Following the reviewer's suggestion, we have further refined the paragraph regarding the EU Circular Economy Action Plan and the German KrWG. We have made it more concise and directly linked it to our quantitative results, highlighting that the observed reduction in consumables and the shift toward digital equipment provide a data-driven basis for waste-prevention strategies and clinical procurement guidelines.

 

Comment: Finally, the conclusions align with the results and limitations; it is good that you explicitly acknowledge potential trade-offs in toxicity/resource depletion for electronics and recommend multi-criteria LCA as future work.  Overall, the revised manuscript is nearing completion for publication; however, it would benefit from (i) a clearer and more audit-friendly presentation of the allocation parameters, (ii) a concise sensitivity analysis regarding hazardous-waste modelling, and (iii) the inclusion of explicit numeric uncertainty intervals within the Results section.    

• Response: Thank you for the final positive evaluation. Regarding the three minor points for further clarification, we have ensured the following in the revised manuscript:

Allocation Parameters: We have ensured that all parameters used for the device-energy allocation (such as lifespan and utilization assumptions) are explicitly described within the Materials and Methods section (Section 2.2).

Sensitivity Analysis: The implications of waste modeling, specifically the role of hazardous waste and different treatment scenarios, are now discussed in more detail in the Discussion (Section 4.1).

Numeric Uncertainty: See license limitation. I only have access tot he Monte Carlo results I already calculated, not for any new calculations.

Author Response File: Author Response.docx