Life-cycle sustainability assessment of public general hospital buildings in China Authors: Kaiyue Huang, Wenjuan Zhang, Xinyue Miao, Tongyu Xu Published: 2026 ABSTRACT AbstractUnder the backdrop of China’s “dual carbon” goals and the high-quality development of healthcare, general hospitals—characterized by high energy consumption and emissions—require systematic and quantitative evaluation to support sustainable transformation. Based on the Life Cycle Sustainability Assessment (LCSA) framework, this study develops a comprehensive evaluation model integrating Environmental Life Cycle Assessment (ELCA), Life Cycle Cost (LCC), and Social Life Cycle Assessment (SLCA). Using “per bed” as the functional unit, an empirical analysis was conducted on 16 public general hospitals across four representative climate zones in China. Through multi-source data collection, questionnaire surveys, and expert weighting, indicators including carbon emissions during construction and operation, lifecycle cost, and patient satisfaction were standardized and integrated into a comparable LCSA index. The results suggest that: (1) climate zone was associated with differences in CO₂e-based environmental performance, with higher per-bed carbon emissions observed in colder regions; (2) operation and maintenance account for 55–60% of total life-cycle cost, and intelligent energy management combined with renewable energy can simultaneously reduce LCC and ELCA; and (3) social performance is generally moderate, with building comfort, spatial organization, and doctor–patient interaction identified as key determinants of patient satisfaction. Significant differences in sustainability performance were observed among cases, with newly built or green-renovated hospitals equipped with intelligent management systems performing better. The findings highlight that LCSA reflects the coupling relationships among environmental, economic, and social dimensions, where improvements in one dimension may generate synergies or trade-offs with others. The proposed framework provides a quantitative and stakeholder-oriented decision-support basis for hospital administrators, architects, and policymakers to promote low-carbon, cost-effective, and human-centered hospital building transformation.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1038/s41598-026-58511-0.Keywords: Public hospital buildings, Life-cycle sustainability assessment, Environmental performance, Life-cycle cost, Patient satisfactionSubject terms: Climate sciences, Environmental sciences, Environmental social sciences, Environmental studies Introduction Hospitals play a crucial role in promoting overall sustainable development and the transition to a circular economy1. Their significance goes beyond merely providing medical services. As resource and energy-intensive institutions, hospitals consume large amounts of water, electricity, and various materials during operation, while generating waste and pollution emissions, which impose significant pressure on the environment2. Globally, the healthcare system has become the second-largest source of greenhouse gas emissions in the building environment sector, after the industrial sector3. Taking the United States as an example, its healthcare system contributes approximately one quarter of global medical-related greenhouse gas emissions4. Between 2010 and 2018, the emissions of the U.S. healthcare system continued to rise, with the carbon dioxide equivalent increasing from approximately 520 million tons to 554 million tons, and the per capita emission level also significantly increased5. Similar trends can be observed in other countries: the Australian healthcare sector accounts for approximately 7% of carbon emissions in the construction sector6, and the Chinese healthcare industry contributed 2.7% of the country’s total greenhouse gas emissions in 20127. Specifically at the operational level of hospitals, high-energy-consuming functional units such as operating rooms have an annual carbon footprint of several thousand tons of carbon dioxide equivalent, and the environmental impact of a single surgery is also highly significant8. Meanwhile, the global healthcare system is facing multiple challenges such as resource shortages, supply chain fluctuations, an aging population, and an increased burden of environmental-related diseases. The continuous growth in medical demand has made hospitals face more complex trade-offs between ensuring service quality and controlling operational costs9. How to reduce the environmental footprint and enhance system resilience while ensuring medical safety and service efficiency has become an important issue that needs to be addressed urgently in the field of healthcare architecture and health policies. Therefore, it is necessary to re-examine the sustainable development path of hospitals from a systemic perspective, and through scientific quantification and comprehensive management methods, achieve the coordination and unity of medical services and ecological responsibility. In recent years, research on building a sustainable healthcare system based on the three-dimensional goals of economy, society and environment has gradually increased10. However, existing studies often focus on a single dimension or specific issues, such as energy efficiency improvement, medical waste management, embodied carbon accounting or green building design, lacking a comprehensive framework that covers the entire life cycle and multiple stakeholders11–14. Moreover, some studies focus on individual technologies or operational improvement strategies, making it difficult to reveal the coupling relationship of hospitals as a complex system at different stages and functions15. Therefore, constructing a systematic assessment tool from an overall perspective is of great significance for supporting long-term strategic decisions and optimizing resource allocation. Life cycle sustainability assessment (LCSA) provides a theoretical foundation and methodological approach for achieving comprehensive sustainable evaluation of hospitals16. Based on the traditional life cycle assessment, LCSA integrates environmental impact assessment, social impact assessment, and life cycle cost analysis, enabling a systematic measurement of the hospital’s overall performance during the planning, design, construction, operation, and renewal stages17–19. At the environmental level, LCSA helps identify high-emission processes and optimize energy structure and material selection20; at the economic level, it can support fiscal sustainability decisions through cost-benefit analysis21; and at the social level, it can assess factors such as medical service equity, patient satisfaction, and employee well-being22. The three aspects are interdependent and together constitute the core framework for the sustainable development of hospitals. It is worth noting that the development of digital and intelligent technologies has provided new opportunities for the sustainable transformation of healthcare23. Innovative approaches such as telemedicine, Internet of Things monitoring systems, minimally invasive surgical techniques, and 3D-printed personalized medications have the potential to enhance service accessibility and quality while reducing resource waste and unnecessary transportation emissions24–26. Information technology and artificial intelligence can also support data-driven energy management and supply chain optimization, driving the healthcare system towards greater efficiency and lower carbon emissions27,28. However, the long-term sustainable benefits of these technologies still require systematic assessment on a life cycle scale. Furthermore, the challenges faced by countries at different development stages in achieving sustainable medical development vary significantly29,30. While developed countries have relatively complete infrastructure, their medical systems are large in scale and have a high total emission31,32; developing countries, although their overall emissions are relatively smaller, may exhibit higher environmental intensity due to lower technical efficiency33,34. Therefore, when conducting hospital sustainability research in a global context, regional differences and institutional environments should be fully considered to form an adaptable evaluation framework. In the context of the “dual carbon” goal and the “Healthy China” strategy being implemented simultaneously, Chinese general hospitals are confronted with a series of context-specific challenges during their sustainable development transformation. Firstly, due to the large population and uneven distribution of regional medical resources, high-quality medical resources are highly concentrated in tertiary general hospitals in big cities, which are constantly operating at an overloaded state, resulting in significantly higher energy consumption per unit area, equipment operation intensity, and the usage of disposable consumables than the design standards35,36. Under this situation, hospitals often prioritize the efficiency of diagnosis and treatment and service capacity, while insufficient investment is made in energy management optimization and carbon emission control37,38. Secondly, most Chinese general hospitals adopt large-scale space organization in the form of “medical streets” and high-density outpatient layouts, with complex personnel flow and crowded waiting spaces. The operation time and load fluctuations of air conditioning, ventilation, and elevator systems are long, making energy consumption and carbon emission management during the building operation stage even more difficult39,40. Moreover, hospital expansions and renovations are frequent, but in the process of updating existing buildings, the substitution of green building materials, control of embodied carbon, and life cycle cost analysis have not yet formed institutional mechanisms, resulting in a lack of systematic connection between the construction stage and the operation stage41. Additionally, from a social perspective, there is tension between the pressure of medical insurance payment, control of patient medical costs, and the demand for high-quality services, making hospitals face uncertainties in terms of funds and return periods when promoting green procurement, the utilization of renewable energy, or intelligent energy-saving systems42. Finally, compared to developed countries, China is still in the process of improving aspects such as the refined classification of medical waste, carbon accounting of supply chains, and data transparency, and lacks a unified sustainable performance evaluation system and standardized indicators43. The combination of these structural and institutional factors has led to unique and complex sustainable transformation challenges for Chinese general hospitals in achieving the coordinated development of environmental emission reduction, economic feasibility, and social equity. In conclusion, although the concept of sustainable development has gradually been integrated into medical management and the practice of green hospital construction, comprehensive sustainable assessment studies for complete hospital buildings and their entire life cycle are still relatively scarce. This situation is particularly prominent in China. Especially in terms of the systematic framework that integrates environmental, economic, and social goals in a unified manner, there is still a lack of a mature and unified methodological system. Therefore, it is necessary to introduce the LCSA concept into hospital design and operation decisions, and construct a multi-dimensional and full-process evaluation model to provide scientific support for the low-carbon transformation and high-quality development of medical buildings. Compared with existing hospital sustainability assessment models, the proposed framework has several methodological distinctions. First, rather than evaluating a single issue such as operational energy use, medical waste, embodied carbon, indoor environmental quality, or green-building compliance, it integrates CO₂e-based ELCA, building-only LCC, and patient-centered SLCA into a unified life-cycle sustainability index. Second, unlike conventional building sustainability assessments that commonly use floor area as the functional unit, this study adopts “per bed” as the primary functional unit because hospital buildings are service-capacity-oriented facilities. A per-m² sensitivity analysis is further conducted to examine whether the results are robust when the evaluation basis shifts from service capacity to building-area efficiency. Third, the framework combines multi-source operational data, building material data, cost data, patient-satisfaction survey data, and expert-derived weights, thereby linking objective environmental and economic performance with perceived social performance. Fourth, by applying the framework to 16 public general hospitals across four representative climate zones in China, the study provides a context-specific and comparable assessment tool for identifying not only sustainability rankings but also the synergies and trade-offs among environmental, economic, and social dimensions. Methods This study adopts the three core dimensions of sustainable development - environment, economy and society—as the analytical framework to conduct a life cycle sustainability assessment (LCSA) for public hospitals. The aim is to provide a quantitative basis for strategic decisions and resource allocation in the medical industry. The study refers to the comprehensive sustainability evaluation idea proposed by Kloepffer and defines LCSA as the integration result of environmental life cycle assessment (ELCA), life cycle cost (LCC), and social life cycle assessment (SLCA)44,. Its basic expression is as follows: This integration framework provides a theoretical basis for the systematic coupling of indicators from different dimensions, and also clarifies the technical approach of this research. Figure 1 shows the overall methodological process of this study, covering several sequential key steps. Firstly, through a systematic literature review, the progress of LCSA-related research in the medical field is sorted out, and the research questions, analytical boundaries, and evaluation scope are clarified. Based on this, the functional units are determined to ensure the comparability of results among different hospitals. At the same time, a questionnaire is designed around the social dimension to obtain the data required for SLCA. Subsequently, during the data acquisition stage, information such as environmental emissions, resource consumption, and cost composition was collected through the operational and financial data provided by the hospital management department, as well as through field research and on-site interviews. This information was used to conduct ELCA and LCC analyses. After completing the impact assessment of each dimension, the results were quantitatively calculated and standardized to eliminate dimensional differences, ultimately forming comprehensive LCSA indicator values that can be used for horizontal comparison. Research flowchart. Overall, the methodological framework of this study is based on multi-source data collection. Through cross-dimensional integration and standardized analysis, a comprehensive evaluation framework capable of systematically reflecting the sustainable performance of public hospitals has been constructed. The subsequent sections will provide detailed explanations of the specific implementation steps and technical details for each dimension. All methods used in this study, including the collection of data via surveys and interviews with hospital personnel, were carried out in accordance with relevant guidelines and regulations of [College of Urban Construction, Changchun University of Architecture]. The experimental protocols, including the survey questionnaires and interview procedures, were reviewed and approved by the [College of Urban Construction of Changchun University of Architecture -Research Ethics Committee], Proposal Number: (2025-02-18-01). Informed consent was obtained from all subjects and/or their legal guardian(s) in the study. Definition of research scope and objectives This study aims to provide evidence-based support for the planning and construction of hospitals and medical facilities in China by integrating sustainable concepts at the design and operation levels, thereby reducing resource consumption intensity and minimizing environmental burdens. The research subjects are limited to public hospitals in China, and a LCSA is conducted within this scope. The system boundary covers the entire life cycle process of the buildings, from raw material extraction and production, construction, operation and maintenance, to demolition and scrapping, comprehensively evaluating the performance in terms of environmental, economic and social dimensions45. Table 1 shows the core evaluation index system of LCSA selected in this study. In life cycle assessment, accurately defining the functional unit is a prerequisite for conducting scientific comparisons among different buildings46. In existing studies, building life cycle analysis usually takes the unit building area as the functional unit to enhance the comparability among different types of buildings47. However, considering that the core function of a hospital lies in providing medical services rather than merely providing building space, this study believes that using area as the sole measurement criterion is insufficient to accurately reflect its service attributes. Based on the results of expert interviews, several experts in medical management and construction fields pointed out that “per bed” should be used as a more representative functional unit, as the number of beds directly relates to the scale of patients that the hospital can accommodate, and can better reflect its service capacity. During the research process, it was also found that there were significant differences in the average floor area per bed among different hospitals (Table 2). If the unit area is used as the comparison benchmark, it may lead to deviations in the evaluation results. For example, hospitals with larger ward areas may have their unit area impact values diluted under the condition of higher total resource consumption, thereby underestimating the actual environmental and cost burden. Therefore, this study ultimately selected “per bed” as the functional unit to ensure that the evaluation results more accurately reflect the service efficiency and sustainable performance of the hospital. This is also in line with previous studies16. Meanwhile, to test the robustness of the results, this study further conducted a functional-unit sensitivity analysis using “per m2” as an alternative functional unit. The sensitivity analysis was used to examine whether the main conclusions would be fundamentally changed when the evaluation basis shifted from service capacity to building area efficiency. Table 1Sustainable assessment indicators for the entire lifecycle of hospital buildings.LCSAELCALCCSLCALife-cycle greenhouse gas emissionsConstruction costComfort levelEnergy costDistributionMaintenance costHumanizationOperation cost Sustainable assessment indicators for the entire lifecycle of hospital buildings. Table 2Basic information of research cases.CaseC1C2C3C4C5C6C7C8Building area (104 m2)7.47.024.26109.410.57.569.9Number of beds10758004501300100017508901100Area per bed (m2)6987947794608590Climate zoneSevere cold areaCold areaAverage window-to-wall ratio0.280.350.420.430.350.380.420.45Building orientationNSNSNSNSNSNSNSNSWall typeReinforced concrete frame combined with various blocksWall thickness (mm)500500500500400450400400Wall insulationPerlite insulation board (100 mm)Hard polyurethane foam insulation board (80 mm)CaseC9C10C11C12C13C14C15C16Building area (104 m2)6.328.183.777.72.620.78.8418Number of beds89093065095050011509501500Area per bed (m2)718858815218093120Climate zoneHot-summer and cold-winter areaHot-summer and warm-winter areaAverage window-to-wall ratio0.450.510.520.580.580.620.590.61Building orientationNSEWEWNSEWNSEWEWWall typeReinforced concrete frame combined with various blocksWall thickness (mm)350300350300300300300300Wall insulationHard polyurethane foam insulation board (40 mm)Rock wool insulation board (60 mm)NS, Northeast-Southwest direction; EW, East-west direction. Basic information of research cases. NS, Northeast-Southwest direction; EW, East-west direction. Case selection and sample representativeness The case selection followed a stratified purposive case-study design rather than a random national sampling design. The primary stratification criterion was the building climate zone, because climate conditions directly influence heating, cooling, ventilation demand, building-envelope requirements, and operational energy consumption in hospital buildings. According to the Chinese building climate zoning standard, China includes five major building climate zones: severe cold, cold, hot-summer and cold-winter, hot-summer and warm-winter, and mild zones (Fig. 2). During the field investigation, the research team attempted to collect complete data from all major climate zones. However, complete and valid life-cycle data could not be obtained for hospitals in the mild climate zone because of data-access restrictions and incomplete historical construction records. Therefore, the final sample included four representative climate zones, with four public general hospitals selected from each zone, resulting in 16 case hospitals. Climate zoning map of Chinese buildings. (Source: Drawn by the authors. ArcGIS 10.8.1 https://www.arcgis.net.cn). Within each climate zone, hospitals were selected according to four criteria: (1) being a public general hospital, to ensure comparability in institutional function, ownership, and management system; (2) availability of relatively complete operational, cost, and building-related data required for ELCA and LCC; (3) location in major cities with mature public medical systems and fully functional general hospitals; and (4) variation in building scale, bed capacity, area per bed, building-envelope characteristics, building age, renovation status, and energy-management configuration. This sampling strategy was intended to support cross-case and cross-climate comparison under controlled institutional conditions. Therefore, the sample can represent typical data-rich public general hospital buildings in major Chinese cities across four climate zones, but it should not be interpreted as statistically representative of all Chinese hospitals. In addition, C12 and C16 are newly-built comprehensive hospitals that have been in operation for the past 5 years. Although they differ in size, both incorporated sustainable development concepts during the design stage and adopted a number of green technologies and management measures, such as energy management systems, efficient water recycling systems, solar photovoltaic systems, heat pump systems, and the application of environmentally friendly materials. At the initial stage of the research, it was expected that these two hospitals might perform better in the LCSA evaluation. However, considering the differences in their scale, functional organization, and technical implementation paths, it is still necessary to objectively compare and conduct in-depth analysis of their comprehensive sustainable performance through systematic evaluation methods. Indicators and data collection CO2e-based environmental life-cycle assessment This study employed the OneClick LCA software to conduct an ELCA for hospital buildings. This tool, based on the lifecycle assessment methodology framework, systematically quantifies the environmental impact of the buildings throughout their entire life cycle and provides standardized result expressions and comparison methods, thereby enhancing the transparency and communicability of the assessment results48. To obtain the necessary basic data, the research team conducted on-site investigations of all case public hospitals. The investigation found that some hospital buildings were constructed earlier and the related technical archives were incomplete. Due to the government’s data confidentiality policy, researchers in some cases were unable to obtain the original construction drawings and engineering quantity lists of the buildings. However, with the approval of the relevant authorities, the research team was able to conduct mapping work at the hospital sites, estimating the main material usage during the construction stage by measuring the building area and combining the construction methods on-site. In terms of operational data, the hospital management provided annual electricity and fuel consumption data as well as some self-generated electricity data with photovoltaic systems, which were used to calculate the environmental impact of operational energy usage. Additionally, the study collected and organized information on the types of building materials, structural forms, energy usage, waste generation volume, and water consumption of each hospital. These data were uniformly entered into the OneClick LCA system to calculate the ELCA results for each case hospital and conduct quantitative analysis using carbon dioxide equivalent CO₂e as the main environmental impact indicator49. Social life-cycle assessment SLCA aims to identify and analyze the positive and negative impacts that products or facilities have on the social aspect throughout their entire life cycle50. In the medical context, medical quality is typically defined as the effectiveness of implementing intervention measures while meeting established safety standards, as well as the ability to achieve health improvements under resource constraints51. In recent years, patient satisfaction has become an important dimension for measuring the performance of hospital services. Relevant studies have pointed out that in situations with higher quality of interpersonal communication such as doctor-patient interaction, the level of patient satisfaction is usually higher52. Empirical research in The United States also indicates that patient satisfaction is one of the core indicators for evaluating hospital nursing quality53. Moreover, when evaluating the indoor environmental quality of hospital buildings, the actual needs and expectations of the users for indoor environmental conditions should be fully considered54. Based on the above theoretical framework, this study conducted questionnaire surveys and semi-structured interviews through a combination of online and offline methods to obtain data related to SLCA. The initial questionnaire items were developed based on a systematic literature review, expert consultation, and semi-structured interviews with hospital users. The item pool was mapped onto the three SLCA dimensions: comfort55, spatial distribution56, and humanized features57, as shown in Appendix A (abc), in this study, these three dimensions are referred to as the secondary indicators. Where, the comfort dimension includes five sub-indicators: acoustic environment, light environment, indoor air quality, ward area environment, and thermal comfort level; the spatial distribution dimension covers spatial flexibility, accessibility and path organization, department layout, and overall well-being support; the humanization dimension is measured through three sub-indicators: safety and security, health promotion, and social interaction. Through this multi-dimensional indicator system, the comprehensive performance of hospital buildings in the social aspect is systematically evaluated16, these sub-indicators are referred to as the tertiary indicators. To obtain the specific data of various indicators and sub-indicators, this study constructed survey questions based on the respondents’ satisfaction, and on this basis identified and quantified the key determinants. For example, under the acoustic environment dimension, it was further divided into four core factors: noise level, sound privacy, the communication experience of medical staff, and the quality of the broadcasting system. By breaking down the macro indicators step by step into measurable specific variables, the study can more systematically identify the key influencing paths of the social sustainability of medical buildings and analyze its promoting effect on the overall well-being of individuals in different environments. This refined measurement method helps to reveal the complex mechanisms behind social sustainability and provides a basis for the formulation of improvement strategies58. This study aims to assess the current social sustainability status of existing hospitals, with the survey subjects being the patient population receiving medical services locally. Using a simple random sampling method, with an error range of 3% and a confidence level of 95%, the required effective sample size was determined to be 830. The questionnaire was measured using a seven-point Likert scale, where 1 indicates complete dissatisfaction, 2 indicates general dissatisfaction, 3 indicates slight dissatisfaction, 4 indicates neither dissatisfaction nor satisfaction, 5 indicates slightly satisfied, 6 indicates generally satisfied, 7 indicates complete satisfaction59. The survey results were used to calculate the satisfaction scores for each key determinant. The data were collected through face-to-face interviews, and a total of 830 valid questionnaires were obtained. Researchers assisted respondents on site to reduce misunderstanding and improve data quality. To ensure the comparability of SLCA scores among the 16 case hospitals, a hospital-stratified balanced sampling strategy was adopted. Each hospital was treated as an independent sampling stratum, and patients were randomly invited to participate in the survey within each hospital. Because the objective of the SLCA survey was to compare hospital-level patient satisfaction scores, the valid questionnaires were distributed as evenly as possible across the 16 hospitals rather than strictly in proportion to hospital bed capacity. Finally, 830 valid questionnaires were obtained, with 51–52 valid responses from each hospital. At the climate-zone level, 208 valid questionnaires were collected from the severe cold area, 208 from the cold area, 207 from the hot-summer and cold-winter area, and 207 from the hot-summer and warm-winter area. This balanced distribution ensured that each case hospital had more than 50 valid responses, which was considered sufficient for calculating hospital-level aggregated SLCA scores and supporting descriptive horizontal comparison among cases. To ensure the measurement quality of the SLCA questionnaire, reliability and validity tests were conducted using SPSS before calculating the final SLCA scores. The questionnaire items were developed by converting the key determinants listed in Table 3 into measurable patient satisfaction questions. In total, the SLCA questionnaire contained 51 items, including 22 items for comfort, 17 items for spatial distribution, and 12 items for humanized features. Table 3Boundary of building-only LCC in this study.Cost categoryIncluded in this study?TreatmentPlanning and design costYesIncluded as initial building-related investmentConstruction and installation costYesIncluded and converted to the same price base yearBuilding material costYesIncluded in construction-stage costBuilding operation energy costYesIncluded, such as electricity, heating, cooling and fuel related to building operationWater and utility cost related to building operationYesIncluded where data were availableRoutine building maintenance costYesIncluded in annual O&M costRenovation and replacement of building systemsYesIncluded in maintenance/renewal costLand acquisition costNoExcluded because public hospitals are commonly built on government-allocated landMedical equipment purchases and maintenanceNoExcluded because this study focuses on the building entity rather than clinical equipmentStaff salariesNoExcluded because they belong to hospital service operation rather than building LCCDrug costs and medical consumablesNoExcluded because they belong to clinical service costsSalvage valueNo, base caseNot separately calculated due to high uncertainty and the public-infrastructure nature of hospitalsDemolition costNo, base caseNot separately calculated because public hospitals are more commonly renovated or expanded than fully demolished Boundary of building-only LCC in this study. The internal consistency of the questionnaire was examined using Cronbach’s alpha coefficient. Reliability analysis was conducted for the overall SLCA scale, the three secondary dimensions, and the twelve tertiary indicators. A Cronbach’s alpha value greater than 0.70 was considered acceptable for internal consistency. The validity of the SLCA questionnaire was evaluated using the Kaiser–Meyer–Olkin (KMO) test, Bartlett’s test of sphericity, and exploratory factor analysis. The KMO test and Bartlett’s test were used to determine whether the questionnaire data were suitable for factor analysis. Exploratory factor analysis was then conducted in two steps. First, the 51 questionnaire items were analyzed to examine whether they corresponded to the twelve tertiary indicators in the SLCA framework. Second, the aggregated scores of the twelve tertiary indicators were further analyzed to examine whether they corresponded to the three secondary dimensions: comfort, spatial distribution, and humanized features. Factors with eigenvalues greater than 1 were retained, and factor loadings greater than 0.50 were regarded as acceptable. After the reliability and validity of the scale were confirmed, the satisfaction scores were used for the weighted aggregation of SLCA. After obtaining the satisfaction scores, the research entered the stage of determining the weights to support the comprehensive calculation of LCSA. For this purpose, another expert questionnaire was designed, targeting professionals with rich practical experience and in-depth understanding of sustainable development, including industry experts, scholars, and hospital staff. The questionnaire used a five-point Likert scale, where 1 indicated not important and 5 indicated extremely important59. The experts were required to rank the importance of the three levels and the key determinants. Due to the fact that the research on social sustainability in the medical field is still in its early stages, the number of expert responses was relatively limited. In the end, a total of 80 valid questionnaires were collected. All participants had many years of experience in their respective fields, making the results highly professional and reliable. In this study, the weights of the first-level dimensions (ELCA, LCC and SLCA), the secondary indicators and the tertiary indicators, as well as the key determinants under each tertiary indicator, were all determined through the Relative Importance Index (RII) method as shown in Formula (2) : where, S represents the score given by the respondents for a certain factor (ranging from 1 to 5), H represents the highest score value (5), and N represents the number of participants in the assessment (the number of experts’ assessment scores is 80). After obtaining the RII values for the indicators, normalization is performed to obtain the weights. The weight normalization follows formula (3). where, represents the normalized weight of the j-th indicator, and represents the relative importance index of the j-th indicator. m represents the number of indicators involved in the weight calculation. The specific quantity in this study is determined according to the number of corresponding indicators at different levels. To evaluate the reliability of the expert-derived weighting scheme, two validation procedures were conducted. First, Kendall’s coefficient of concordance (W) was used to examine the degree of agreement among the 80 experts regarding the relative importance of the indicators. Because the expert questionnaire used a five-point Likert scale, the original importance scores were converted into within-level ranks before calculating Kendall’s W. The test was conducted separately for the first-level LCSA dimensions, the secondary SLCA indicators, the tertiary SLCA indicators (Formula 4–5). Second, the dispersion of expert scores was examined using the coefficient of variation (Formula 6). where, is the sum of ranks assigned to indicator j, is the mean rank sum, N is the number of experts, and m is the number of indicators being weighted at the same level. where, is the coefficient of variation of expert scores for indicator j, is the standard deviation of expert scores, and is the mean expert score. Lower CV values indicate lower dispersion and more stable expert judgments. Formula (7) is used to aggregate the scores of the key determinants to obtain the score of the tertiary indicators. The tertiary indicator scores are then weighted by the weights of the tertiary indicators to obtain the secondary indicator scores, and these secondary indicators are further aggregated to obtain the SLCA value for each hospital: where, represents the weight of key determinant i, represents the score of the corresponding key determinant, and represents the comprehensive score of indicator j. Building-only life-cycle cost In this study, LCC refers specifically to the building-only life cycle cost of public general hospital buildings, rather than the total cost of the hospital system. The assessment object is the hospital building entity and its building-related operation and maintenance activities. Therefore, the LCC boundary includes planning and design costs, construction and installation costs, building material costs, building-related energy and utility costs during operation, routine maintenance costs, and renovation or replacement costs of building systems. Costs directly associated with clinical service provision, including medical equipment procurement and maintenance, staff salaries, drug costs, and medical consumables, were excluded. Land acquisition costs were also excluded because most public hospitals in China are built on government-allocated land. This boundary definition is consistent with the objective of this study, which is to compare the economic sustainability of hospital buildings rather than the total financial performance of hospital institutions. A 50-year reference study period was adopted for all case hospitals. This period was used as a standardized comparative time horizon rather than as an assumption that every hospital building would be demolished after 50 years60. The use of a common 50-year period allows hospitals built in different years and located in different climate zones to be compared under the same temporal boundary. Because the 16 cases are all permanent public hospital buildings, the 50-year reference period was considered appropriate for evaluating the long-term economic burden of building construction, operation, maintenance, and renewal. To ensure comparability among hospitals built in different years, all cost data were converted to constant prices in the same base year before discounting. In this study, [2024] was selected as the price base year. Construction-stage costs recorded in the year of construction were adjusted to the base year using construction cost indices61. Operation and maintenance costs were adjusted to the same base year using consumer price indices, utility price indices, or local cost information where applicable62. Therefore, the LCC results were calculated using constant-price cash flows rather than nominal cash flows63. The historical costs are unified to the same base period price as shown in Formula 8. where, represents the amount of the k-th type of cost of hospital h converted to the base price period; represents the nominal amount of this cost in the year y; represents the corresponding price index in the base price period; represents the corresponding price index in the year of cost occurrence. The net present value method was used to aggregate costs over the 50-year reference period. A real discount rate of 3.5% was adopted in the base-case calculation to reflect the long-term time value of capital in public building investment. Because all costs were converted to constant prices, no additional inflation rate was applied in the NPV calculation. Regional cost differences were retained as part of the actual local economic burden of hospital building operation. The NPV of the building-only LCC is shown in Formula (9). The bed allocation is standardized as shown in Formula (10). where, represents the net present value of the lifecycle cost of the building entity of hospital h; represents the initial construction-related costs in the base period; represents the building operation energy and utility costs in the t-th year; represents the building operation maintenance costs in the t-th year; represents the building renovation and replacement or system replacement costs in the t-th year; r is the discount rate, and the base scenario in this article takes 3.5%. represents the number of hospital beds, and represents the lifecycle cost of the building structure per bed, with the unit being RMB/bed. Inflation was handled by converting all cost data to constant prices in the same base year before discounting. Therefore, the NPV calculation was conducted using real cash flows and a real discount rate, rather than nominal cash flows with an additional inflation rate. Regional cost differences were not completely removed because the purpose of this study was to evaluate the actual economic burden of hospital buildings under their local context. However, all costs were converted to the same price base year and then expressed per bed to improve cross-case comparability. Considering that public hospitals are large-scale public infrastructure assets and are usually renovated, expanded, or functionally upgraded rather than fully dismantled at the end of a fixed service period, residual value and demolition costs were not separately included in the base-case calculation64. Instead, renovation and replacement activities during the reference period were included in the maintenance and renewal cost category. This treatment was adopted as a boundary assumption of the building-only LCC and is further discussed in the limitations section. Table 3 presents the LCC boundary used in this study. Comprehensive evaluation The final stage of this study involves the comprehensive calculation of the LCSA values. Since LCC, ELCA, and SLCA are presented in different units and dimensions (such as monetary values, CO₂e, and satisfaction scores), they need to be standardized before integration and converted to the same functional unit basis. Since ELCA and LCC are cost-related indicators, the lower the value, the better the performance. Therefore, reverse min-max standardization is adopted for them. SLCA is a benefit-related indicator, and the higher the value, the better the social performance. Therefore, positive min-max standardization is used for it. This study selects “impact per bed” as the functional unit to enhance the comparability among hospitals of different sizes. Specifically, the results of ELCA and LCC are first converted into “per bed” indicators, while SLCA, as a dimensionless hospital-level score based on patient satisfaction, is no longer divided by the number of beds. The normalization of positive indicators is as shown in Formula (11). Reverse indicators need to be positively transformed, as shown in Formula (12). where, represents the normalized value of the positive indicator for a certain case. represents the normalized value of the negative indicator for a certain case. represents the value of the positive indicator. represents the value of the negative indicator. represents the maximum value of this indicator. represents the minimum value of this indicator. After normalization, the results of the three pillars are weighted and integrated to form a single LCSA comprehensive value for each hospital. The determination of weights is based on the results of expert questionnaires and interviews. Experts ranked and scored the importance of the three dimensions of sustainability, and then weighted and adjusted the standardized data accordingly. In the context of Chinese public hospital buildings, experts generally believe that the importance of LCC is relatively low. This judgment is mainly based on the following real background: Firstly, Chinese public hospitals are public welfare infrastructure, with significant social service attributes. Their construction and operation are centered on meeting public health needs rather than maximizing economic profits. Secondly, hospitals are rigid public service facilities. Even if the operating costs are high, they are difficult to be shut down or closed due to economic reasons. Therefore, under the framework of policy guidance and social responsibility, cost control alone is not the primary evaluation criterion. In contrast, experts generally emphasize the crucial role of the environmental and social dimensions in hospital architecture in China. On one hand, with the advancement of the “dual carbon” goals and the implementation of green hospital construction policies, hospital buildings, as high-energy-consuming public buildings, have received increasing attention for their environmental performance (such as energy efficiency and carbon emission control); on the other hand, public hospitals bear the social responsibility of ensuring the health of the entire population. Social factors such as patient experience, indoor environmental quality, and the environment for doctor-patient interaction directly affect the quality of medical services and public satisfaction. Therefore, in the comprehensive evaluation, the environmental and social weights are relatively higher, which better reflects the practical orientation towards sustainable development of hospital architecture in China. To further reveal the dynamic relationships and potential feedback mechanisms among the three pillars, this study constructed a causal loop diagram (Causal Loop Diagram, CLD) based on the system thinking approach to identify the interdependent relationships among the various indicators of LCSA and the possible reinforcing or balancing mechanisms they may form16. This system model helps to deeply understand the complex coupling relationships among the economic, environmental and social dimensions of hospital buildings throughout their entire life cycle, and provides theoretical support for formulating comprehensive optimization strategies. The reliability verification of expert ratings, standardized weights, and weights based on expert opinions is shown in Tables 4 and 5, and the final comprehensive scores for each case are presented in Formula (13). Table 4Expert scoring and weights of the first-level dimensions.DimensionThe total score of 80 expertsRII valueNormalized weightsELCA3520.88000.383LCC2050.51250.223SLCA3630.90750.394 Expert scoring and weights of the first-level dimensions. Table 5Reliability validation of expert-derived weights.Weighting levelNumber of expertsNumber of indicatorsKendall’s Wχ2dfp-valueCV rangeMean CVFirst-level LCSA dimensions8030.46273.922< 0.0010.126–0.2470.170SLCA secondary indicators8030.38661.762< 0.0010.152–0.2180.184SLCA tertiary indicators80120.298262.2411< 0.0010.166–0.2910.228 Reliability validation of expert-derived weights. where, y represents the comprehensive score of the case. represents the normalized value of each indicator. The Kendall’s W test showed that the expert judgments had statistically significant concordance at each weighting level. This indicates that the expert-derived weights were not randomly distributed and that the RII-based weighting results had acceptable inter-rater consistency. The coefficient of variation of the expert importance score is within a reasonable range, further indicating that the dispersion of expert judgments was within an acceptable range. Therefore, the expert-derived weights were considered sufficiently reliable for the subsequent LCSA aggregation. Causal loop diagram construction Causal loop diagrams (CLDs) are useful tools for interpreting the causal relationships among variables in a complex system16. In this study, the CLD was developed to show how the environmental life cycle assessment (ELCA), life cycle cost (LCC), and social life cycle assessment (SLCA) indicators interact within the life cycle of Chinese public general hospital buildings. The variables were selected from the main indicators and empirical findings of the manuscript, especially carbon emissions per bed, area per bed, operation and maintenance cost, smart energy management, renewable energy use, indoor environmental quality, spatial organization, and patient satisfaction. Results CO2e-based ELCA results This study uses carbon dioxide equivalent CO₂e as a unified evaluation indicator for the environmental dimension, integrating emissions from the production of building materials, the construction stage, daily operation, and maintenance process to form a single quantitative result. The analysis shows (see Fig. 3), the average emission level of all case hospitals is 8873.5 kg/bed/year. Comparisons by climate zones reveal significant differences: the average value is the highest in the extremely cold region, at 11,518 kg/bed/year; in the cold region, it is 8639 kg/bed/year; in the hot summer and cold winter region, it is 7213 kg/bed/year, which is the best performance; in the hot summer and warm winter region, it is 8124 kg/bed/year. The overall trend indicates that climate conditions have a significant impact on the energy consumption and emissions during the hospital operation stage. Regions with higher heating loads have relatively higher carbon emission levels. At the specific case level, C12 and C16 are newly built hospitals that have been in use for the past five years. Their overall environmental performance is better than that of most existing hospitals, which is closely related to their adoption of low-carbon technologies and smart energy management systems during the design and operation stages. Meanwhile, both C3 and C16 used low-carbon and environmentally friendly materials during construction, may have contributed to embodied carbon emissions. Although C3 in the extremely cold region is an existing hospital, it has introduced a smart energy management system in recent years, and performs relatively well among the four cases in the same region. However, due to the high heating demand brought by the extremely cold climate, its overall emission level is still higher than that of C12 and C16 in the warm climate region. Comparing C12 and C16, C16 has a larger single-bed floor area (120 m2/bed), which theoretically is not conducive to controlling emissions per unit bed. However, C16 has a larger site and roof area, which can accommodate a larger photovoltaic system, thereby increasing the proportion of self-generated electricity and offset some external power consumption. Additionally, its outpatient hall adopts a passive ventilation design, may have contributed to cooling energy consumption in large spaces. Therefore, although C16 does not have an advantage in the single-bed area indicator, it still achieved relatively excellent environmental performance, Although C16 did not have an advantage in the area-per-bed indicator, it still achieved relatively favorable CO2e-based ELCA performance. This may be associated with its larger available roof area for photovoltaic installation and the use of passive ventilation design in the outpatient hall. CO2e-based ELCA results of the 16 case hospitals. In contrast, C14 in the hot summer and warm winter region performed the weakest among the environmental indicators in this climate zone. Its single-bed floor area is as large as 180 m2/bed, the largest among all cases, making it difficult to achieve an advantage under the evaluation system based on “per bed”. However, this hospital is a teaching and research-oriented hospital with significant influence in the region, equipped with complete and large-scale teaching and research facilities. This case demonstrates that as a complex functional complex integrating medical care, hospitals integrating medical care, teaching, and research have distinctive resource-input patterns. A single environmental indicator is difficult to fully reflect its comprehensive value, and it is necessary to conduct multi-dimensional sustainable comprehensive evaluations. In the comparison of internal conditions in cold regions, the emission performance of C1 was slightly better than that of C2 and C4. Within the severe-cold-region cases, C1 had a lower window-to-wall ratio and slightly lower emissions than C2 and C4. This pattern may suggest a possible association between envelope design and heating-related energy demand. Although the lower window-to-wall ratio weakened the natural lighting conditions to some extent, it significantly improved the insulation performance of the building envelope, reduced the winter heating energy consumption, and thus improved the overall environmental performance. This result suggests that building-envelope design may be important for energy saving and emission reduction in energy conservation and emission reduction in hospitals in cold climate areas. Both the cases in the cold regions and the cold climate areas adopted a north-south orientation layout, which was more conducive to providing heat from the winter sunlight, while the cases in hot summer cold regions and hot summer warm regions mostly adopted an east-west orientation layout, which was more conducive to shading in summer and reducing the cooling demand. SLCA results Before comparing the SLCA scores among the 16 case hospitals, the reliability and validity of the patient satisfaction questionnaire were examined. The results showed that the SLCA questionnaire had satisfactory internal consistency. The Cronbach’s alpha coefficient of the overall SLCA scale was 0.952, indicating excellent reliability. The Cronbach’s alpha values of the three secondary dimensions were 0.928 for comfort, 0.914 for spatial distribution, and 0.887 for humanized features. At the tertiary-indicator level, the Cronbach’s alpha values ranged from 0.781 to 0.874, indicating acceptable to good reliability for all subscales. In addition, the corrected item-total correlation coefficients ranged from 0.463 to 0.781, and deleting any item did not substantially improve the reliability of the corresponding scale. Therefore, all questionnaire items were retained for subsequent analysis. The validity test results also supported the suitability of the questionnaire for factor analysis. The KMO value of the overall SLCA questionnaire was 0.941, and Bartlett’s test of sphericity was significant, χ2 = 23846.572, df = 1275, p < 0.001. This indicates that the correlations among the items were sufficient for factor extraction. The KMO values of comfort, spatial distribution, and humanized features were 0.921, 0.904, and 0.873, respectively, and Bartlett’s tests for the three dimensions were all significant at p < 0.001. Exploratory factor analysis was first conducted at the item level. Twelve factors with eigenvalues greater than 1 were extracted, corresponding to the twelve tertiary indicators in the SLCA framework. The cumulative variance explained was 68.74%, and the factor loadings of all retained items ranged from 0.612 to 0.857. A further factor analysis based on the twelve tertiary-indicator scores extracted three factors, corresponding to comfort, spatial distribution, and humanized features. These three factors explained 67.86% of the total variance. The rotated factor loading matrix showed that all tertiary indicators loaded most strongly on their theoretically assigned dimensions, with primary factor loadings ranging from 0.741 to 0.846. These results indicate that the SLCA questionnaire had satisfactory reliability and construct validity and was suitable for calculating patient-satisfaction-based social performance scores. The results of the reliability and validity tests for the quantitative table are shown in Tables 6, 7 and 8. Table 6Reliability and validity test results of the SLCA questionnaire.Scale/dimensionNumber of itemsCronbach’s alphaKMOBartlett’s test of sphericityOverall SLCA scale510.9520.941χ2 = 23846.572, df = 1275, p < 0.001Comfort220.9280.921χ2 = 7964.318, df = 231, p < 0.001Spatial distribution170.9140.904χ2 = 5682.447, df = 136, p < 0.001Humanized features120.8870.873χ2 = 3269.581, df = 66, p < 0.001 Reliability and validity test results of the SLCA questionnaire. Table 7Reliability results of the tertiary SLCA indicators.Secondary dimensionTertiary indicatorNumber of itemsCronbach’s alphaComfortAcoustic environment40.831ComfortLight environment50.862ComfortIndoor air quality40.846ComfortWard area environment50.874ComfortThermal comfort level40.819Spatial distributionSpatial flexibility50.852Spatial distributionAccessibility and path organization40.824Spatial distributionDepartment layout30.793Spatial distributionOverall well-being support50.866Humanized featuresSafety and security50.841Humanized featuresHealth promotion40.828Humanized featuresSocial interaction30.781 Reliability results of the tertiary SLCA indicators. Table 8Rotated factor loading matrix based on the twelve tertiary indicators.Tertiary indicatorFactor 1: ComfortFactor 2: Spatial distributionFactor 3: Humanized featuresAcoustic environment0.8120.2140.168Light environment0.8460.1870.152Indoor air quality0.8230.2050.191Ward area environment0.7910.2360.174Thermal comfort level0.7760.1910.208Spatial flexibility0.2260.7980.194Accessibility and path organization0.1830.8170.176Department layout0.2450.7620.213Overall well-being support0.1980.7850.246Safety and security0.1720.2140.741Health promotion0.2030.2310.803Social interaction0.1960.2560.782 Rotated factor loading matrix based on the twelve tertiary indicators. The eigenvalues of the three extracted factors were 4.382, 2.274, and 1.487, respectively, and the corresponding variance explained was 36.52%, 18.95%, and 12.39%. The cumulative variance explained by the three factors was 67.86%. After confirming the reliability and construct validity of the questionnaire, the SLCA scores of the 16 case hospitals were calculated based on patient satisfaction. The SLCA evaluation covered three secondary dimensions: comfort, spatial distribution, and humanized features. As illustrated in Fig. 4, most cases scored between 4 and 5 points overall, corresponding to a neutral to slightly satisfied level. This indicates that patients generally held moderate perceptions of hospital performance and that substantial room for improvement remains in the social dimension. The SLCA score of the case. Case C12 achieved the highest overall score of 6.13, reaching a clear satisfaction level. This result may be associated with its relatively recent construction, compact spatial organization, and better-rated indoor environmental conditions, resulting in high ratings for lighting conditions, acoustic environment, indoor air quality, and thermal comfort. In contrast, C1, C2, C6, and C9 performed relatively poorly in terms of comfort. These hospitals are older buildings located in colder climate regions, where design limitations and climatic constraints make it difficult to incorporate features such as atriums to enhance daylighting and environmental quality. Moreover, their outdated design concepts no longer align with contemporary expectations for healthcare comfort. In terms of spatial distribution, location plays a significant role in shaping patient evaluations. Although C12 is smaller in scale, it is situated in the city center, benefiting from convenient accessibility and compact spatial organization. Its spatial distribution score is approximately 10% higher than that of C16. However, due to limited parking availability in central urban areas, C12 scored about 15% lower than C16 on the “private transportation” indicator. C16, a larger suburban hospital, benefits from more extensive green spaces, parks, and parking facilities, which enhance its performance in transportation convenience and environmental openness. These findings highlight the contrasting advantages and disadvantages between centrally located hospitals and large suburban medical complexes. Among hospitals in cold climate regions, C3 is the only case with an overall score exceeding 5. C3 is the only case in the severe-cold region with an overall SLCA score above 5. This may be related to its recent renovation and improved indoor environmental quality. Its recent comprehensive renovation has significantly improved the indoor environmental quality. The introduction of the intelligent energy management system not only enhanced the environmental performance but also may have had a positive impact on the social evaluation results. Owing to its central urban location, C3 performs well in accessibility and circulation organization. However, its score in the “social interaction” sub-indicator under humanized features is more than 20% lower than that of C11, C12, C14, and C16, suggesting the need for improved communication and interaction between medical staff and patients. Notably, the “safety and security” indicator shows consistently high satisfaction across all 16 cases, with minimal variation. Patients generally assume hospitals to be safe environments, resulting in relatively low sensitivity to this factor. Case C11, a public specialty hospital with strong national recognition, achieved a relatively high overall SLCA score despite lower comfort ratings due to its older facilities. This outcome is largely attributable to its effective management system, particularly in the human-centered dimension, where patient satisfaction exceeds the sample average by approximately 17%. Case C14, the largest hospital in the sample, has been in operation for many years but was originally planned as a national demonstration hospital. Benefiting from its high-quality initial design and sustained management optimization, it performs well across all three dimensions—comfort, spatial distribution, and humanized features. Case C1 received the lowest overall score, primarily due to aging infrastructure and inefficient spatial organization, resulting in a comfort score approximately 13% below the sample average. Although its central urban location provides advantages in accessibility and circulation, limited green space and insufficient parking reduce its spatial distribution rating. Furthermore, patient feedback indicates deficiencies in staff–patient communication, leading to a human-centered score approximately 23% below the average. Overall, the SLCA results demonstrate that building age, climatic conditions, geographic location, and management quality collectively influence the social performance of hospitals. Beyond physical infrastructure, continuous renovation and effective service management are critical factors in enhancing patient satisfaction and improving the social sustainability of healthcare facilities. Building-only LCC results Figure 5 presents the LCC results for all cases, revealing substantial variation among hospitals. The LCC value consists of two major components: initial construction investment and subsequent operation and maintenance (O&M) costs. Differences in architectural standards—particularly the significant variation in floor area per bed—contribute to disparities in life-cycle cost, as larger per-bed areas generally lead to higher operational expenditures. Among all cases, C3, C12, and C16 are the only hospitals equipped with smart energy management systems, and they demonstrate comparatively lower LCC values. Detailed analysis shows that approximately 55%–60% of total life-cycle cost (after discounting) are concentrated in the operation and maintenance phase. This finding indicates that the traditional approach of “prioritizing construction over operation” is economically inefficient in the long term. As public hospitals, management teams often focus primarily on improving medical service quality, while lifecycle operational efficiency receives comparatively less attention, leading to overlooked cost-saving opportunities. Building-only LCC results of the 16 case hospitals. In cold regions, C3 is the only major hospital that has recently upgraded to adopt an intelligent energy management system. This might enable it to demonstrate superior economic performance compared to other hospitals in the same climate zone. Among all the samples, C12 has the lowest LCC value. Its relatively compact building size, smaller per-bed area, combined with efficient design and management practices, may help minimize the life cycle cost. C16 also performs quite well; however, its per-bed area is larger, resulting in its LCC being slightly higher than that of C12. By contrast, C8 exhibits the highest LCC value. This hospital has been in operation for a long period, and both its building envelope and energy plant equipment are outdated. Field investigations reveal that hospitals with higher LCC values generally suffer from low-efficiency HVAC systems operating continuously without intelligent controls or manual optimization. Lighting systems also demonstrate considerable inefficiencies. In some cases, windows are obstructed by trees or blinds, limiting daylight penetration and leading to excessive reliance on artificial lighting without smart regulation. Behavioral and operational factors further exacerbate energy waste. In severe cold and cold regions, main entrances are frequently left open during winter, causing significant heat loss. In hot summer–cold winter and hot summer–warm winter regions, doors and windows are often opened while air-conditioning systems are running, as occupants seek fresh air, thereby increasing cooling loads. In contrast, C12 and C16 are equipped with advanced mechanical ventilation systems that ensure adequate indoor air quality. This reduces the need to open windows during HVAC operation and enhances overall energy efficiency, contributing to their relatively favorable LCC performance. Comprehensive LCSA results To improve the transparency of the comprehensive evaluation, Table 9 reports the complete calculation process of the LCSA score for the 16 case hospitals. The table includes the original values of ELCA, LCC, and SLCA, their standardized values, the weighted contribution of each first-level dimension, and the final LCSA score. This makes the weighted aggregation process explicit and allows the results shown in Fig. 6 to be reproduced. Table 9Transparent calculation process of the LCSA score for the 16 case hospitals.CaseELCA rawELCA std.ELCA contributionLCC rawLCC std.LCC contributionSLCA rawSLCA std.SLCA contributionLCSAC111,5470.2520.096135,3340.4260.0953.790.0000.0000.186C213,3400.0000.000157,3830.1830.0414.080.1240.0490.086C385430.6730.258112,0790.6830.1525.140.5770.2270.637C412,6430.0980.037161,0760.1420.0324.430.2740.1080.177C597320.5060.194137,0980.4070.0914.790.4270.1680.455C676750.7950.304125,4090.5360.1194.250.1970.0770.504C783100.7060.270156,9470.1880.0424.980.5090.2000.523C888410.6310.242173,9390.0000.0004.870.4620.1820.437C975990.8060.309143,9560.3310.0744.230.1880.0740.464C1083670.6980.267150,8090.2550.0574.560.3290.1300.462C1169540.8960.343121,6450.5770.1295.340.6620.2610.739C1262131.0000.38383,3671.0000.2236.131.0000.3941.000C1368950.9040.346113,5410.6670.1494.980.5090.2000.699C1410,4780.4020.154149,0730.2750.0615.790.8550.3370.557C1583280.7030.269134,8480.4320.0964.670.3760.1480.518C1667980.9180.352100,5370.8100.1815.780.8500.3350.870ELCA raw values are expressed as kg CO₂e/bed/year; LCC raw values are expressed as NPV per bed, RMB/bed; SLCA raw values are patient satisfaction scores. ELCA and LCC were standardized using reverse min–max normalization, while SLCA was standardized using positive min–max normalization. The first-level weights were 0.383 for ELCA, 0.223 for LCC, and 0.394 for SLCA. Transparent calculation process of the LCSA score for the 16 case hospitals. ELCA raw values are expressed as kg CO₂e/bed/year; LCC raw values are expressed as NPV per bed, RMB/bed; SLCA raw values are patient satisfaction scores. ELCA and LCC were standardized using reverse min–max normalization, while SLCA was standardized using positive min–max normalization. The first-level weights were 0.383 for ELCA, 0.223 for LCC, and 0.394 for SLCA. Comprehensive LCSA scores of the 16 case hospitals. The comprehensive LCSA scores for each case are weighted and integrated based on the three-dimensional indicators of environment, society and economy (LCC). To enhance the comparability among different hospitals, the three indicators are first standardized, and then the weights are determined through expert consultation, ensuring that the final scores fall within the range of 0–1. Figure 6 shows the comprehensive evaluation results for each case. Overall, C12 achieved the highest score, indicating that it holds significant benchmarking significance within the sample of this study. In comparison, the other hospitals showed certain gaps in different dimensions, suggesting that there is still considerable room for improvement. Analyzing by climate zone, the combined scores of the severely cold regions and the cold regions were generally low. Only C3 performed relatively well after completing building renovations and introducing a smart energy management system, achieving a high score of 0.637; the scores of the other cases were all below 0.6, indicating that achieving comprehensive sustainable improvement still faces significant challenges in high heating load climate conditions. C16 also adopted advanced low-carbon design strategies and a relatively complete management system, and its overall performance was at a relatively good level. However, due to its larger building scale and higher unit bed area, under the evaluation system where the functional unit is “per bed”, its overall score was slightly lower than that of C12. In contrast, Case C2 had the lowest comprehensive score, indicating that it ranked at the bottom in terms of environmental performance, social satisfaction, and lifecycle cost control. It urgently needs systematic improvements in multiple aspects such as building renovation, operation management, and service quality. Table 10 presents the sensitivity analysis results of LCSA after the conversion of units. Table 10Sensitivity analysis of LCSA results under the per-m2 functional unit.CaseArea per bed, m2/bedOriginal LCSA score, per bedRank, per bedSensitivity LCSA score, per m2Rank, per m2C1690.186140.03716C2870.086160.15914C3940.63750.6554C4770.177150.13415C5940.455120.5086C6600.50490.22913C7850.52370.4977C8900.437130.4628C9710.464100.30812C10880.462110.4579C11580.73930.43910C12811.00010.8983C13520.69940.32111C141800.55760.9372C15930.51880.5385C161200.87020.9391 Sensitivity analysis of LCSA results under the per-m2 functional unit. The sensitivity analysis showed that the main conclusions were generally robust, but the relative performance of some large hospitals changed when the functional unit was changed from “per bed” to “per m2”. Under the original per-bed functional unit, C12 ranked first and C16 ranked second, while C14 ranked sixth. Under the per-m2 functional unit, C16 ranked first, C14 ranked second, and C12 remained in the top three. This indicates that C12 and C16 are robust high-performing cases, while the performance of C14 is sensitive to the functional unit because of its large teaching and research-related floor area. These results do not invalidate the use of “per bed” as the primary functional unit, but they indicate that “per bed” reflects service-capacity-based sustainability performance rather than pure building-area efficiency. For hospitals with large teaching, research, and regional medical center functions, per-bed results should be interpreted with caution because this functional unit may amplify the apparent environmental and cost burden of non-ward functional spaces. The case of C14 illustrates the importance of functional-unit selection. Under the per-bed functional unit, its large teaching and research-related floor area increases the apparent carbon and cost burden per bed. However, the per-m2 sensitivity analysis shows that C14 performs much better from the perspective of building-area efficiency. Therefore, the relatively lower per-bed score of C14 should not be interpreted as poor overall building performance, but rather as a result of its complex functional composition. Causal loop analysis In the ideal sustainability scenario, the hospital building system is expected to reduce carbon emissions and other environmental impacts, control life cycle cost, and improve patient-centered social performance at the same time. The selection of key indicators is based on the high weight scores given by experts in the survey within this study. In the comprehensive life cycle assessment, bed capacity and service demand, as well as climate heating/cooling loads, are regarded as the main driving factors. Bed capacity affects the size of the building, while climatic conditions increase the heating or cooling loads. These driving factors influence construction costs, operation and maintenance costs, operating energy consumption, and carbon emissions. Figure 7 presents the interdependencies among these variables and the three sustainability dimensions. A minus sign means that an increase in one variable decreases the connected variable, or that a decrease in one variable increases the connected variable. The CLD contains six feedback loops: four reinforcing loops (R1, R2, R3, and R4) and two balancing loops (B1 and B2). Reinforcing loops amplify change and may either improve system performance or deepen an undesirable lock-in. Balancing loops counteract change and help the system move toward equilibrium. The causal cycle diagram of systems thinking. R1, known as the cost loop, begins with the bed capacity. Higher service demands often lead to an increase in the building size and the area per bed, which will raise construction costs, operational energy consumption, and maintenance costs. When the building area increases, it will also intensify waste generation, thereby further increasing the R2 Emission Loop. This cycle explains why an overly large area per bed may impose long-term carbon emissions and cost burdens on the hospital. B1, known as the conservation loop, describes the enhanced effects resulting from building insulation and intelligent energy operations. Intelligent energy management can improve operational energy efficiency, thereby reducing energy consumption and operation and maintenance costs. Lower operating and maintenance costs lead to cost savings, even though this may increase the initial investment. However, in the long run, the benefits outweigh the investment. This cycle reflects the findings of this paper, namely that operations and maintenance account for 55% to 60% of the total life cycle cost, and intelligent management is a key intervention point for reducing the total life cycle cost and environmental impact cost. B2 is called the resource efficiency loop and is a balancing mechanism. It demonstrates the impact of water resource collection, sunlight utilization, and material recycling on building costs. These technologies can reduce construction costs and operating costs. At the same time, it is connected to the R2 Emission Loop, reflecting the interdependence among ELCA indicators. This cycle is consistent with the performance of cases such as C12 and C16, in which green design concepts, photovoltaic systems, passive ventilation, and efficient management help to offset some environmental burdens. The satisfaction cycle R3 encompasses the key factors for enhancing social sustainability. It brings about a series of impacts, including comfort, health promotion, air quality, spatial flexibility, and social influence. The higher the comfort level, the more human interventions it represents and thereby affects the R2 emission loop. R4, the distribution loop, promotes social sustainability. It interfaces with R1 at the operational and maintenance cost variables and is connected to R3 at the “social aspect” node. It depicts a relationship: an increase in safety will increase the operational and maintenance costs of the hospital, although it is not ideal, but at the same time, it enhances the hospital’s social aspect, which is a positive effect. Discussion Climate-related carbon burden in hospital buildings The first issue emerging from the results is the clear climate-related difference in CO₂e-based ELCA performance. In this study, hospitals located in severe cold and cold regions showed higher carbon emissions per bed than those in hot-summer and cold-winter regions. This pattern is consistent with previous studies on hospital and public-building energy use, which have emphasized the dominant role of operational energy demand, especially heating, cooling, and ventilation loads, in shaping building-related carbon emissions65,66. In hospital buildings, this effect can be further amplified because medical facilities normally operate for long hours and require stable indoor temperature, ventilation, hygiene, and safety conditions. This finding also extends the existing literature by showing that climate is not merely a background variable but a boundary condition for hospital sustainability assessment. Under the same functional unit of “per bed”, hospitals in colder regions may face a structural disadvantage because heating demand increases the operational carbon burden. Therefore, the relatively weak ELCA performance of severe-cold-region hospitals should not be interpreted only as a failure of design or management. Rather, it reflects the interaction between climatic constraints, building envelope performance, operational intensity, and energy system efficiency67. At the case level, C3 provides a useful example. Although it is located in the severe cold region, its CO₂e-based ELCA performance was better than that of other cases in the same climate zone, possibly because of recent renovation and the introduction of smart energy management. This suggests that climate-related disadvantages may be partly mitigated by envelope improvement, operational control, and energy management, although the present 16-case sample does not allow a causal conclusion. In this sense, the results are consistent with the broader argument in the building LCA literature that operational-stage intervention remains a critical pathway for reducing the carbon burden of existing public buildings68. Intelligent energy management as a possible ELCA–LCC synergy The second issue concerns the relationship between environmental and economic performance69,70. However, the results of this study suggest that, in hospital buildings, intelligent energy management may create a possible synergy between ELCA and LCC during the operation stage. This interpretation is supported by two observations. First, the LCC results showed that operation and maintenance accounted for a large proportion of total life cycle cost. Second, cases equipped with smart energy management systems, such as C3, C12, and C16, tended to perform better in LCC and also showed relatively favorable CO₂e-based ELCA performance. This pattern is consistent with previous studies suggesting that digital technologies, Internet of Things monitoring, and artificial-intelligence-supported energy management can improve operational efficiency and reduce resource waste in healthcare systems71,72. It also aligns with the LCSA perspective that environmental, economic, and social dimensions are interdependent rather than isolated evaluation pillars73. The possible mechanism is that smart energy management reduces unnecessary operation of HVAC, lighting, and mechanical ventilation systems. This can lower electricity and fuel consumption, thereby reducing CO2e emissions, while also reducing long-term operation and maintenance costs. In this sense, intelligent management is not simply an environmental technology; it can also act as an economic management tool. The cases of C12 and C16 suggest that when energy monitoring, equipment maintenance, renewable energy use, and management incentives are combined, environmental and economic performance may improve simultaneously. Trade-offs within SLCA: comfort, spatial accessibility, and humanized features The third issue concerns the internal structure of the social dimension. The SLCA results show that social sustainability in hospital buildings cannot be reduced to a single satisfaction indicator. Instead, it involves interactions and possible trade-offs among comfort, spatial accessibility, and humanized features. This is consistent with the patient-satisfaction and hospital indoor-environment literature, which emphasizes that patient experience is shaped by both physical environmental quality and interpersonal or institutional factors74,75. The comparison between C12 and C16 illustrates this trade-off. C12, located in a central urban area, benefited from compact spatial organization and good accessibility, which may have contributed to its strong spatial distribution score. However, its limited parking conditions weakened its private-transportation-related evaluation. In contrast, C16, as a large suburban hospital, had more space for green areas, parking, and environmental openness, but its peripheral location may reduce accessibility for some patients. This contrast suggests that central hospitals and suburban medical complexes may have different social-performance advantages: the former may provide better urban accessibility, while the latter may provide better spatial openness and environmental capacity. A second trade-off concerns comfort and environmental-economic efficiency. Larger floor area, more open public space, atriums, and green areas may improve perceived comfort and spatial experience, but they may also increase floor area per bed, operational energy demand, and maintenance cost. Therefore, improving comfort does not automatically lead to better overall sustainability if it produces excessive carbon and cost burdens. Conversely, overly compact hospital layouts may improve ELCA and LCC performance but may reduce patient comfort, wayfinding quality, waiting-space experience, and staff–patient interaction. A third issue is that humanized features may partly compensate for physical limitations. For example, some older hospitals did not perform strongly in comfort because of aging facilities, but their humanized-feature scores could remain relatively favorable when management quality, staff–patient communication, and safety perception were strong. This suggests that the social sustainability of hospital buildings depends not only on architectural form but also on service culture and institutional management. Therefore, SLCA should be understood as a socio-spatial assessment rather than a purely physical-space assessment76. Overall, the SLCA findings suggest that social sustainability involves a balance among spatial quality, accessibility, comfort, and humanized service. The goal is not to maximize one dimension in isolation, but to identify combinations that improve patient experience without generating excessive environmental or economic burdens. Implications for sustainable hospital building transformation The above discussion has two implications for sustainable hospital building transformation. First, LCSA is useful because it reveals whether an intervention produces synergy or trade-off across dimensions. For example, intelligent energy management may improve both ELCA and LCC, while excessive expansion of floor area may improve comfort but increase carbon emissions and cost. Second, hospital sustainability strategies should be differentiated according to climate zone, building age, and functional type. For hospitals in cold regions, envelope renovation and heating-system optimization should be prioritized. For large teaching and research hospitals, the functional-unit choice should be interpreted carefully because non-ward spaces may increase per-bed environmental and cost burdens. For hospitals with weak SLCA performance, physical renovation should be combined with improvements in wayfinding, accessibility, service communication, and management culture. Therefore, the contribution of this study is not only to rank the 16 cases, but also to show how hospital sustainability assessment can be used to identify cross-dimensional relationships. In this sense, the LCSA framework provides a basis for moving from single-indicator optimization to integrated decision-making in hospital building planning, renovation, and operation. Conclusions Based on the theoretical framework of Life Cycle Sustainability Assessment (LCSA), this study developed a hospital-building-specific evaluation model integrating CO₂e-based Environmental Life Cycle Assessment (ELCA), building-only Life Cycle Cost (LCC), and patient-centered Social Life Cycle Assessment (SLCA). Using “per bed” as the functional unit, the framework was applied to 16 public general hospitals across four representative climate zones in China. Through multi-source data collection, patient satisfaction surveys, expert weighting, normalization, and integrated scoring, this study established a quantifiable and comparable method for evaluating the life-cycle sustainability performance of public general hospital buildings. The results show three main findings. First, climate zone was associated with differences in CO₂e-based environmental performance. Hospitals in severe-cold and cold regions generally showed higher per-bed carbon emissions, which may be related to heating demand and climatic constraints. However, cases with low-carbon design strategies, energy-saving renovation, renewable-energy measures, and intelligent energy management tended to show better environmental performance, suggesting that climate disadvantages can be partly mitigated through design and operational optimization. Second, the operation and maintenance stage accounted for approximately 55%–60% of building-only life-cycle cost, indicating that operation-stage management is a key intervention point for improving long-term economic performance. Cases with intelligent energy management, optimized equipment maintenance, and renewable-energy measures tended to have lower LCC values. Third, social performance was generally at a moderate level and was mainly affected by building comfort, spatial organization, accessibility, and doctor–patient interaction. This indicates that improving the social sustainability of hospital buildings requires coordinated enhancement of both the physical environment and institutional service management. The comprehensive LCSA results revealed clear differences among the 16 cases. C12 achieved the highest overall score and can be regarded as a benchmark case with relatively balanced environmental, economic, and social performance. Some hospitals in colder regions showed lower comprehensive scores, suggesting that achieving multi-dimensional sustainability remains challenging under high heating demand and high operational burden. More importantly, the LCSA results indicate that ELCA, LCC, and SLCA are not independent dimensions. Improvements in one dimension may generate synergies or trade-offs with others. Therefore, sustainable hospital transformation should be understood as a multi-objective balancing process rather than the optimization of a single indicator. The findings provide practical implications for hospital administrators, architects, and policymakers. For hospital administrators, the LCSA framework can support operation-stage sustainability management by linking carbon emissions, life-cycle cost, and patient-perceived social performance. Routine monitoring of energy use, CO₂e emissions, maintenance cost, medical waste, and patient satisfaction should be strengthened, and priority should be given to measures such as intelligent energy monitoring, HVAC optimization, preventive maintenance, renewable-energy operation, and improvements to waiting areas, wayfinding, indoor comfort, and accessibility. For architects and design teams, LCSA can be introduced into early planning, new construction, and renovation decision-making. Design alternatives should be compared not only by initial construction cost, but also by life-cycle carbon emissions, long-term operating cost, and user experience. For policymakers, LCSA indicators can be incorporated into green hospital standards, public hospital renovation programs, project approval, and post-occupancy evaluation. Climate-zone-specific benchmarks for hospital carbon intensity and life-cycle cost intensity should also be developed to support more targeted policy guidance and public investment. The main methodological contribution of this study lies in translating the general LCSA concept into a practical assessment procedure for public general hospital buildings. By combining CO2e-based ELCA, building-only LCC, patient-centered SLCA, service-capacity-based functional-unit selection, expert weighting, and cross-dimensional interpretation, the proposed framework provides comparative advantages over single-dimension or rating-based hospital sustainability assessment models. It can support integrated decision-making by identifying not only which hospitals perform better, but also how specific strategies may generate environmental–economic–social synergies or trade-offs. Overall, this study enriches the application of LCSA in the field of medical buildings and provides a quantitative decision-support basis for the low-carbon and high-quality development of public general hospitals in China. Nevertheless, the findings should be interpreted as evidence from 16 data-rich public hospital cases rather than as nationally representative estimates. Future research should expand the sample size, include more hospital types and climate zones, incorporate emissions from medical equipment and supply chains, include healthcare staff and other stakeholder perspectives in SLCA, and conduct long-term dynamic tracking to further improve the sustainability assessment of hospital systems. Limitations and future research First, the generalizability of the findings is limited by the sample size and sampling strategy. This study included 16 public general hospitals across four building climate zones in China. Although the sample was balanced by climate zone and all cases provided relatively complete life-cycle data, it should be interpreted as a climate-zone-stratified case-study sample rather than a statistically representative national sample. Potential sampling bias may arise from four aspects. First, no hospital from the mild climate zone was included because complete and valid data could not be obtained, which limits the climatic completeness of the analysis. Second, all cases were public general hospitals located in major cities, so the findings may not fully represent private hospitals, specialized hospitals, county-level hospitals, or rural hospitals. Third, the requirement for complete operational, cost, and building-related data may have favored hospitals with better data-management capacity, leading to data-availability bias. Fourth, the equal allocation of four cases per climate zone supports balanced comparison but does not correspond to the actual national distribution of hospitals across climate zones. Therefore, the results should be understood as exploratory evidence for typical data-rich urban public general hospitals rather than as nationally generalizable estimates. Future research should expand the sample size, include the mild climate zone, incorporate hospitals of different ownership types, grades, and functions, and use stratified random or multi-center sampling to improve representativeness and enable statistical testing of climate-zone effects. Secondly, in terms of data acquisition, due to the relatively early construction period of some hospitals or the data confidentiality policies involved, some construction material usage and historical technical data could not be fully obtained. In the research, on-site measurement and estimation methods were adopted to make up for the deficiencies, which might to some extent affect the accuracy of the environmental assessment results. Additionally, the SLCA indicators are mainly obtained through patient satisfaction surveys, and their evaluation results are influenced by factors such as sample structure, subjective perception, and short-term medical experience. They have not fully incorporated broader social dimensions such as the working environment of medical staff, management efficiency, and community impact. Future research can combine multi-party survey data, long-term tracking data, and hospital management data to further improve the evaluation system from a social perspective. Thirdly, although “per bed” was selected as the primary functional unit because it reflects the service capacity of hospital buildings, this unit may not fully capture the functional complexity of large teaching, research, or regional medical center hospitals. The per-m2 sensitivity analysis showed that the ranking of some cases, especially C14, changed when the functional unit was shifted from service capacity to building-area efficiency. Therefore, the per-bed results should be interpreted as service-capacity-based sustainability performance rather than as an absolute judgment of overall building efficiency. Future research should further compare multiple functional units, such as per bed, per m², per inpatient bed-day, and per annual patient volume, to better reflect the diverse functional compositions of hospital buildings. Finally, this study mainly employed the static life cycle assessment method to conduct a horizontal comparison of the comprehensive performance of different hospitals at a certain time period. It has not yet covered the long-term dynamic change process. With the development of digital healthcare, smart hospitals, and low-carbon energy systems, the operation mode of hospital buildings is constantly evolving. Future research can combine digital monitoring data, dynamic simulation methods, and system dynamics models to conduct dynamic tracking and analysis of energy consumption, carbon emissions, and social performance of hospital buildings during the long-term operation stage. This will provide more forward-looking decision support for the continuous optimization of medical buildings and policy formulation. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Acknowledgements Project supported by the Key Research and Development Program of Hangzhou, China (Grant No. 2024SZD1A29), Medical and Health Science and Technology Plan Project of Zhejiang, China (Grant No. 2025KY078). Abbreviations Life cycle sustainability assessment Environmental life cycle assessment Life cycle cost Social life cycle assessment Causal loop diagram Author contributions Conceptualization, methodology, software, validation, formal analysis, investigation, resources, K.H.; data curation, writing—original draft preparation, writing—review and editing, W.Z. and T.X.; visualization, supervision, project administration, funding acquisition, X.M. All authors have read and agreed to the published version of the manuscript. Data availability Data are available with the first author and can be shared upon reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Wenjuan Zhang, Email: zwj423@ZJU.edu.cn. Tongyu Xu, Email: xutongyu0409@outlook.com. References Associated Data Supplementary Materials Supplementary Material 1 Data Availability Statement Data are available with the first author and can be shared upon reasonable request.