This investigation uses a retrospective, multicenter observational design to examine ED presentations across multiple health care facilities in the greater Lübeck area and southeastern Schleswig-Holstein, Germany. The study period spans January 1, 2014, to December 31, 2024, encompassing an 11-year observation window. This timeframe was deliberately selected to capture temporal dynamics before, during, and after the COVID-19 pandemic, thereby enabling the assessment of both long-term climatic trends and the impact of infection control measures that may have altered health care–seeking behavior and disease transmission patterns. The extended observation period provides sufficient statistical power to detect seasonal variations, lag effects, and nonlinear exposure-response relationships while accounting for temporal confounding. An overview of the overall study workflow is provided in Figure 1.

This protocol is reported in accordance with the RECORD (Reporting of Studies Conducted Using Observational Routinely-Collected Health Data) statement, an extension of the STROBE (Strengthening the Reporting of Observational Studies in Epidemiology) guideline developed specifically for studies using routinely collected health data [39,40]. A completed RECORD checklist is provided as Checklist 1.

This study incorporates 4 health care facilities representing different levels of care across the region. The University Hospital Schleswig-Holstein (UKSH) Lübeck serves as a regional maximum care provider, contributing data from both the interdisciplinary and pediatric EDs. Sana Kliniken Lübeck is a large urban hospital serving the metropolitan area. Schön Klinik Neustadt provides specialized cardiovascular and orthopedic care in a midsized coastal town. The DRK Hospital Mölln-Ratzeburg serves a rural catchment area in southeastern Schleswig-Holstein. This diversity in care provision levels, geographic distribution, and catchment area characteristics enhances the generalizability of the findings while enabling the examination of potential differences in climate-health relationships across health care delivery contexts and socioeconomic populations.

The primary outcome will be the daily count of ED presentations, stratified by disease category. The secondary outcomes include category-specific presentations of cardiovascular, respiratory, psychiatric, infectious, renal, neurological, and surgical/traumatic conditions. Additional analyses will examine presentations for specific conditions that are particularly relevant to climate and air quality, including asthma and chronic obstructive pulmonary disease exacerbations, acute respiratory infections, acute coronary syndromes, heart failure, heat-related illnesses, and cold-related conditions. Age-stratified analyses will focus on toddlers (<5 y), children/adolescents (5‐18 y), working-age adults (18‐65 y), and older adults (≥65 y) to identify age-specific vulnerabilities.

Based on previous similar investigations [4,6] and preliminary data from the participating centers, we anticipate approximately 800,000 to 1,000,000 ED presentations across the 4 facilities over the 11-year study period. This substantial sample size provides excellent statistical power to detect modest effect sizes, even in subgroup analyses. For example, with daily counts averaging 200 to 250 presentations across facilities, we have >90% power to detect relative risk increases of 5%‐10% per 10 °C temperature change at an α level of .05, consistent with effect sizes reported in comparable studies [5,7,8]. The multicenter design enhances the power to examine heterogeneity, while the extended time series enables robust control for seasonal and long-term trends [34].

Data extraction and harmonization protocols will standardize the information across participating centers, addressing potential differences in coding practices, data completeness, and documentation standards. A centralized database will be established with quality control procedures, including range checks, logical consistency checks, and verification of temporal sequences. Missing data patterns will be systematically characterized, and regular data quality reports will be generated to identify anomalies requiring investigation. All data linkage procedures, including the temporal and spatial matching of environmental exposures to health outcomes, will be documented and validated.

All data will be stored on password-protected local and institutional servers of the University Hospital Lübeck (UKSH Lübeck), with access restricted to authenticated members of the study group. To minimize the risk of reidentification of pseudonymized patient data through cross-referencing covariables such as postal code, age, and sex, individual-level data will not be shared beyond the authorized study team. Results will be released exclusively in aggregated form as daily case counts.

Analytical code will be developed in R and stored locally, with a transfer to a version-controlled repository planned to ensure transparency and reproducibility. Code will be reviewed by an independent member of the study team to minimize the risk of analytical errors. Analysis scripts will be made available upon reasonable request.

Analytical code will be maintained under Git-based version control in a private institutional repository, with access secured via multifactor authentication and regular encrypted backups. Analytical steps will be documented in reproducible notebooks integrating code, results, and narrative, so that tables and figures are generated directly from the analysis without manual transfer. Code will be independently re-executed in a clean computing environment by a study-team member not involved in the original implementation, and identified issues will be tracked and resolved through the repository’s issue-tracking system.

This retrospective study will use pseudonymized routine clinical data without individual patient contact. The legal basis for data processing is established under Section 38 of the State Hospital Act (Landeskrankenhausgesetz, LKHG) of Schleswig-Holstein [47] (effective since July 5, 2024), which permits further processing of patient data for research purposes, provided that the use of these data does not compromise the treatment of patients and that the fundamental principles of the General Data Protection Regulation and the State Data Protection Act are adhered to. Ethical approval was granted by the responsible ethics committee in July 2025.

In accordance with Section 6 of the Health Data Usage Act (Gesundheitsdatennutzungsgesetz; [48]), lawfully stored data may be further processed for medical research purposes, with pseudonymization applied at the earliest possible stage of data processing. Prior to data transfer, all direct identifiers, including patient identification numbers, were removed by the providing institutions.

Ethical approval was obtained from the ethics committee of the University of Lübeck (AZ 2025‐269), including review of the inclusion of vulnerable population groups. The study protocol was reviewed by the institutional data protection officers at all participating facilities. Data-sharing agreements between institutions specify the permitted uses, security requirements, and duration of data retention. Results will be reported in aggregate form only, with measures to prevent the potential identification of individuals in small cells (ie, groups with fewer than 5 cases). The inclusion of vulnerable populations, including children, individuals with chronic conditions, and socioeconomically disadvantaged groups, is scientifically necessary, given their documented elevated risk of weather-related health impacts. Exclusion of these groups would bias the results and potentially perpetuate health inequities by limiting the evidence for targeted interventions.

This study will be conducted in accordance with Article 35 of the Declaration of Helsinki (DRKS-ID: DRKS00037454). The ethics committee waived the need for ethical approval and patient consent for the collection, analysis, and publication of retrospectively obtained anonymized data for this noninterventional study.

The study schedule and the current status of data acquisition and analysis are summarized in Table 1.

Project planning and conceptualization commenced in June 2024.

Clinical data collection was conducted sequentially across the 4 participating hospitals. Data from 3 of the 4 sites were successfully received in July 2025, August 2025, and October 2025, respectively. The fourth and final dataset is anticipated in the second half of 2026. No deviations from the originally planned study protocol have occurred.

Meteorological data from the German Weather Service were obtained in October 2025, covering the full study period from January 2014 to December 2024. Complete air quality measurements from state environmental monitoring were received in January 2026. Infectious disease surveillance data from the Robert Koch Institute were obtained in September 2025.

Across the three datasets received, a preliminary case count indicates that 601,734 ED presentations have been recorded. Baseline characteristics of included patients, including age distribution, sex, and ICD-coded primary diagnoses, will be described in full upon completion of data collection. No major deviations from the planned inclusion criteria have been identified in the datasets received.

Descriptive statistical analyses of the available data have been initiated. Inferential statistical modeling will be performed upon receipt and integration of all data sources.

The study findings will be disseminated through peer-reviewed publications in international journals focusing on environmental health, public health, and emergency medicine. The results will be presented at relevant scientific conferences. A lay summary will be prepared for communication with the participating institutions, public health authorities, and the general public. Individual participating facilities will receive site-specific reports to inform local adaptation plans. Data will be made available through recognized research data repositories following appropriate embargo periods and in compliance with data protection regulations, with individual-level data being pseudonymized and aggregated as necessary.

This investigation has several methodological strengths that enhance the validity and generalizability of the anticipated findings. The multicenter design incorporating four health care facilities across different care provision levels and geographic locations provides diverse patient populations and health care delivery contexts, enhancing external validity beyond single-center studies. The 11-year observation period offers substantial statistical power while enabling robust control of temporal trends, seasonal patterns, and long-term climatic shifts [4,6]. Critically, the study period encompassed the prepandemic, pandemic, and postpandemic phases, permitting the assessment of how infection control measures and altered health care–seeking behaviors may have modified climate-health relationships.

Comprehensive exposure assessment integrates multiple meteorological parameters, air quality measures, and infectious disease surveillance data, enabling the examination of independent, interactive, and synergistic effects [3,4]. The use of DLNM represents a widely adopted analytical framework in environmental epidemiology [14,34], enabling the characterization of both immediate and delayed health impacts while accounting for nonlinear exposure-response relationships [14,18]. The inclusion of facilities serving diverse socioeconomic populations combined with postal code–level geographic information enables the investigation of health inequities in climate-related ED utilization [7,16,26,49], addressing an important gap in the German context. Finally, the focus on northern Germany, with its distinctive maritime climate, contributes evidence from an understudied region where heat-health and cold-health relationships may differ from continental climate zones that have dominated European climate-health research [30-32]. This investigation builds upon prior population-based studies from the Hamburg metropolitan area that have established temperature-dependent patterns for various emergency conditions under marine west coast climate conditions [21,22,25,33], extending this evidence base through multicenter, multifacility analysis of ED presentations.

Several limitations warrant consideration when interpreting the study findings. The retrospective observational design precludes causal inference in the strict experimental sense, although the temporal nature of exposure-outcome relationships and the use of time-series methods with appropriate confounder adjustment provide evidence for causal associations following the Bradford Hill criteria [34]. The ecological approach to socioeconomic characterization using postal code-level data introduces potential ecological fallacy, as within-area heterogeneity cannot be captured, and individual-level socioeconomic circumstances may differ substantially from area-level averages [26].

Weather and air quality exposure assignment based on fixed monitoring stations may introduce measurement errors, particularly for individuals residing at greater distances from monitoring locations or spending substantial time in locations different from their residence [8]. This classical exposure measurement error would typically bias effect estimates toward the null hypothesis, suggesting that true associations may be stronger than those observed. The lack of individual-level data on factors such as housing characteristics, air conditioning availability, occupational exposures, and behavioral adaptations limits the ability to identify specific mechanisms underlying observed associations or to characterize highly vulnerable individuals within broader demographic groups [5,7].

The generalizability of the findings to other German regions or international contexts requires careful consideration, given the potential differences in population characteristics, health care system organization, climate adaptation measures, and baseline climate conditions [31,32]. The focus on ED presentations rather than community-level disease occurrence means that the findings reflect health care utilization conditional on seeking emergency care, potentially missing milder cases managed in primary care or not seeking medical attention. Changes in health care access, insurance coverage, or facility capacity during the study period may have confounded the temporal trends in presentation rates. Finally, while the study examined multiple disease categories, residual confounding by unmeasured factors, including co-occurring environmental exposures, population mobility patterns, or temporal changes in diagnostic coding practices, cannot be entirely excluded despite careful analytical approaches.

This investigation is expected to generate important evidence with implications for public health preparedness, health care system planning, and climate adaptation strategies in northern Germany and comparable maritime climate regions [3]. By quantifying the associations between meteorological factors and ED utilization across multiple disease categories and patient subgroups, the findings will inform the development of targeted early warning systems and preventive interventions [16,50,51]. The identification of vulnerable populations and high-risk periods will enable more efficient allocation of public health resources and clinical capacity during extreme weather events [52,53].

A comprehensive assessment of the impact of air quality alongside meteorological factors will contribute to integrated environmental health strategies addressing both climate change mitigation and adaptation [3,28]. Evidence of synergistic effects between temperature, air pollution, and infectious disease incidence would support the need for coordinated surveillance and intervention programs [4]. The multicenter design encompassing different care provision levels will inform health care system resilience planning, highlighting which facility types and patient populations experience the greatest climate-related pressures.

From a scientific perspective, this study advances methodological approaches to multicenter climate-health research in Germany while contributing to the limited evidence base regarding maritime climate regions in Europe [29,31]. The inclusion of the COVID-19 pandemic period provides unique opportunities to examine how infectious disease epidemics and associated control measures modify climate-health relationships, with potential relevance for future pandemic preparedness. The findings may inform the revision of clinical guidelines for managing heat- and cold-related illnesses [14], the development of facility-level heat action plans [54], and the integration of climate considerations into ED operations and staffing models.

More broadly, this study will support evidence-based climate change adaptation in the health sector, contributing to national and international climate-health assessments [3,32]. By demonstrating the feasibility and value of linking routine health data with environmental monitoring for climate-health surveillance, this study may catalyze similar investigations in other regions and support the establishment of ongoing surveillance systems [16]. The open data sharing approach will facilitate secondary analyses, meta-analyses, and methodological comparisons, advancing the field of climate-health research.

Climate change poses substantial and growing challenges to emergency health care delivery through multiple pathways, including temperature extremes, air pollution, and altered infectious disease patterns [3]. This comprehensive multicenter investigation will characterize these relationships in the context of northern Germany’s distinctive maritime climate, identify vulnerable populations, and quantify health system impacts over an 11-year period spanning the COVID-19 pandemic. By integrating rigorous environmental epidemiology methods [34] with routine health care data across multiple facilities and care provision levels, this study will generate actionable evidence to inform climate adaptation strategies, public health interventions, and health care system preparedness [16,51]. The findings will contribute to protecting population health in the face of accelerating climate change while advancing the scientific understanding of climate-health relationships in understudied European regions [31,32].