Emergent climate protection strategies in German hospitals: A cluster analysis

PLOS One | https://doi.org/10.1371/journal.pone.0312661 May 16, 2025 7 / 18 Results Sample characteristics Out of the 1,169 hospitals surveyed, the final sample comprised 205 respondents, yielding a response rate of 17.5%. Seventy-four respondents (36%) reported holding an administrative leadership position in multiple hospitals, boosting the study’s implications beyond 205 hospitals within Germany. The majority of the respondents were male (81.5%), with a median age ranging between 50 and 59 years. The respondents’ median work experience was 10–14 years (detailed information in S1 Table). Hospital characteristics reveal a median range of 300–499 beds per hospital. Approximately 39.02% of the surveyed hospitals provide primary and standard care. Hospital ownership was evenly distributed. Hospital administrative leaders from single hospitals (40%) primarily participated in the survey (detailed information in S1 Table). Overview Implementation of climate actions Regarding the implementation of climate actions, climate actions in the area of energy were particularly prominent (mv = 5.12). They were most comprehensively implemented in hospital engineering (mv = 5.04), whereas their implemen- tation in medical departments was comparatively limited. Key performance indicators were most frequently collected for energy consumption (mv = 5.38) and water consumption (mv = 5.07). The implementation of climate actions was most comprehensive driven by technical staff (mv = 5.33), while medical staff demonstrated significantly lower levels of engage-

ment (see Tables 1–2). Overall, 27.8% of hospitals have established a sustainability committee, whereas only 14.1% have introduced a cost center. Clustering emergent climate protection strategies in German hospitals Cluster analysis identified four distinct groups of emergent strategies for climate protection in German hospitals. The clusters, each with unique characteristics, are described in Table 3. According to the Bonferonni post-hoc test, all selected cluster vari- ables significantly contribute to cluster separation. The results of the MANOVA indicate statistically significant differences among Table 2. Cluster description: Initiatives and projects. Centroids MANOVA Test Statistics Characteristics (mv) Cluster 1 (32.20%) Cluster 2 (26.34%) Cluster 3 (13.17%) Cluster 4 (28.29%) Total F-Value Sig. 1 Responsible employees a Medical employees 2.65 2.91 3.93 3.80 3.22 11.246 <.001 b Nursing employees 3.14 3.29 4.2 4.43 3.65 13.683 <.001 c Medical-technical employees 3.00 3.42 3.61 4.18 3.49 7.599 <.001 d Functional service employees 2.99 3.29 3.68 4.02 3.43 6.576 <.001 e Administrative employees 4.18 4.66 4.83 5.32 4.7 7.468 <.001 f Technical employees 5.22 5.12 5.29 5.90 5.33 4.159 0.007 2 Prospective responsible employees a Hospital administrative leaders 5.87 6.11 6.34 6.65 6.2 2.917 0.035 b Medical employees 3.50 4.11 4.80 4.98 4.26 6.829 <.001 c Nursing employees 3.52 4.18 4.83 4.95 4.29 7.09 <.001 d Medical-technical employees 3.72 4.27 4.90 5.16 4.43 6.881 <.001

e Functional service employees 3.43 4.11 4.93 4.78 4.22 7.952 <.001 f Administrative employees 4.48 5.19 5.32 5.68 5.12 5.023 0.002 g Technical employees 5.94 6.06 6.12 6.40 6.11 1.136 0.335 https://doi.org/10.1371/journal.pone.0312661.t003