To the best of our knowledge, this is the first study re-assessing IPC structures in a large number of hospitals through the IPCAF. The number of participating German acute care hospitals (n = 660, response rate of 43.1%) was around 10% lower than in 2018 (n = 739, response rate of 50.2%) although the method of distribution (via the German KISS network) and participation (via an online survey) did not change. Given the length of the questionnaire as well as ongoing challenges due to stretched IPC resources in many hospitals, the response rate and overall participation are nevertheless remarkably high and document a good uptake of the IPCAF as an IPC assessment tool.

Overall, the median IPCAF score remained almost unchanged (693 vs. 690) with a similar proportion of hospitals achieving an advanced (86.6% vs. 84.5%), intermediate (13.2% vs. 15.1%) or basic (0.2% vs. 0.4%) IPC category. The high degree of concordance between results from 2018 and 2023 is striking given the COVID-19 pandemic occurred between the two surveys. The pandemic saw an increase in attention to certain aspects of IPC. However, our survey indicates that this may not have resulted in structural changes that could have been appreciated by the IPCAF. Similarly, other studies that have investigated the effects of the pandemic on IPC structures and practices, have revealed that the COVID-19 pandemic may have triggered a shift in the allocation of IPC resources rather than all-encompassing improvements [16, 17]. Of note, it is important to acknowledge that IPC structures as documented in the survey 2018, were already at a high level in Germany before the pandemic, leaving less room for improvement, and that the pandemic may have triggered changes that were not captured by the IPCAF.

The median scores of all CC, except CC5 (multimodal strategies for implementation of IPC interventions), were above 75, which reflects that these are generally well-established in Germany. This finding documents the important role of that IPC plays in the German healthcare system, which can for example be seen in the long history of the widely used national HAI surveillance network “KISS” [18].

In 2018, the lowest median score (75) was reported for CC7 (workload, staffing and bed occupancy) [3]. The results from 2023 show a considerable increase in the median score for this CC, especially for questions pertaining to the assessment of staffing levels, healthcare worker to patient ratios and mechanisms to determine staffing needs and react to changes. This could be interpreted as an effect of novel directives and laws with a focus on staffing in the German healthcare system (e.g. the “Nursing Personnel Strengthening Act” [19]), or as a consequence of an increased awareness for the aspects bed occupancy and staff-to-patient ratio that may have developed during the COVID-19 pandemic [20, 21]. Among other things, the Nursing Personnel Strengthening Act provides for a nursing staff ratio for hospitals, which is intended to adjust the level of nursing staff to the nursing workload. This is designed to set a lower limit for nursing staffing levels, which the hospital may not fall below [19]. The broad range of scores for CC7 achieved by hospitals illustrates that the level, to which aspects of workload and staffing are incorporated into IPC practices may vary considerably across German hospitals. Additionally, it is conceivable that deficits observed in the IPCAF 2018 concerning staffing, have motivated some hospitals to better address this aspect.

In 2018, the median score of CC5 (80) was comparatively low, which might have been attributable to the relative novelty of the concept of multimodal strategies for IPC interventions, and potential unfamiliarity of respondents with the concept. One could expect that five years later familiarity with this concept should have been higher and that multimodal strategies would have become more integrated in IPC practices. However, our survey revealed scores for CC5 that were slightly lower than in 2018. Correspondingly, the majority of publications on national IPCAF surveys also identified CC5 as a component with comparatively low scores [5, 7, 10, 11, 13]. It appears that despite the resources and literature already provided on the matter [1, 14, 22, 23], the concept and benefits of multimodal strategies in IPC may still not be fully appreciated by currently implemented IPC standards. While unfamiliarity with the concept of multimodal strategies might be an explanation for the seeming non-progression of multimodal strategies in German IPC practices, it can also be speculated that there could be structural or organizational barriers that render the adoption of multimodal strategies difficult. Such potential underlying barriers should be explored in more focused studies, for instance by means of qualitative surveys on the status quo of multimodal strategies for the implementation of IPC in German hospitals. However, as for CC7, the range of scores for CC5 was particularly broad. This indicates that multimodal strategies may be routinely used by some hospitals, while not being employed or only being employed at a rudimentary level by others, thus documenting variation in how IPC activities are implemented across German hospitals. Further aspects that were identified and discussed in our publication on the IPCAF 2018 as potential targets for improvement (e.g. definition of objectives for IPC programs, application of interactive methods to perform IPC training and feedback of surveillance data, integration of IPC aspects into the training of other specialties [3]), showed no meaningful changes, which documents a continued potential for improvement in certain areas of IPC.

Only a minority of hospitals (38.5%) reported having implemented IPC training for patients or family members (CC3). This corresponds to various reports on the matter that showed that IPC education for patients is not yet widely implemented in many hospitals [24]. Given reports of improved patient hand hygiene through education and subsequent lower infection rates [25,26,27], it seems appropriate to suggest that patient-directed IPC training programs should be promoted more widely in German hospitals.

Overall, our study demonstrated that usage of the IPCAF on a broad scale is feasible, and that conclusions can be carefully drawn from its repeated application. However, to fully assess the many aspects of IPC a multitude of IPC instruments are necessary. The WHO offers additional tools like the hand hygiene self-assessment framework [28] or assessment tools on infection prevention and control minimum requirements for health care facilities [29, 30]. It is important to see the IPCAF as part of a growing set of global IPC instruments that can be utilized synergistically.

Our study had several limitations. First, data from the participating hospitals constitute a convenience sample and cannot be seen as representative for Germany. All included datasets were from hospitals participating in the German national surveillance network, which may have a greater than average interest in matters of surveillance and IPC. However, due to the high number of participating hospitals (around one third of hospitals listed in the German hospital register [15]) careful extrapolations to the national level appear justifiable. Second, as per agreement with the study participants, data of the survey were not linked at the level of individual hospitals, to other surveillance data or to data from the 2018 IPCAF survey. Accordingly, observed differences in IPCAF scores could be attributable to a different cohort of participating hospitals rather than actual changes. This reduces the precision of longitudinal analyses and should be addressed with a revised approach in possible future surveys. Third, certain concepts addressed in the IPCAF (e.g. multimodal strategies) are rather complex and might not always have been completely understood. This was addressed by several footnotes and explanations throughout the tool. Forth, some questions may have been perceived as potentially compromising and, despite the confidential nature of the survey, might have prompted wishful reporting. Fifth, after submission of responses to the NRC, participants could not retroactively correct errors or otherwise modify the entered data.