Abstract

Methods

Since 1997, the German national Hospital Infection Surveillance System (KISS) has served as a procedure to monitor nosocomial infections in Germany (7). Within the KISS framework, the MRSA-KISS module was established as a pathogen surveillance component in 2006 (8, 9). As with the other KISS modules, participation in the MRSA component is voluntary. This module can be used for collecting data on multidrug-resistant organisms, as required by section 23 of the German Protection against Infection Act (IfSG). The MRSA module collects information on MRSA detection among hospitalized patients for the entire hospital; rehabilitation hospitals are also eligible to participate in MRSA surveillance. Data collection and transmission to the National Reference Center (NRZ, Nationales Referenzzentrum) for Surveillance of Nosocomial Infections is performed via the online platform “webKess“ (9, 10).

The definitions used for data collection are as follows (8, 9): A patient in whom MRSA is detected in screening swabs (colonization) or isolated from clinical material (colonization or infection) is considered an MRSA case. Hospital cases, not patients with MRSA are counted, meaning that if a patient in whom MRSA was detected is readmitted after discharge from hospital, they are considered a new case. With regard to hospital cases overall (with and without detection of MRSA), we will use the term patients in the following for ease of understanding; however, the hospital admissions of the patients are counted (9). Each case of MRSA is only once considered for the hospital, meaning that if the patient is transferred to another department of the same hospital, the case is not counted again.

The number of MRSA hospital days is calculated from the day of diagnosis or detection of MRSA (date of swab or specimen collection) or from the day of admission, if MRSA had already been detected before hospital admission (9):

• until the day of discharge of the MRSA case (the day of discharge is counted)

• until the patient has no detectable MRSA colonization – as determined by the participating hospitals—for example, after collection of three MRSA-negative screening specimens on three consecutive days.

Participating hospitals are asked to submit once a year the following data on MRSA cases as well as hospital structural and process parameters:

• Number of patients admitted to the hospital per calendar years (number of cases)

• Patient days for a calendar year

• Rehabilitation facility available (yes/no)

• Number of nasal swabs (adjusted for patients yes/no, i.e. only one swab is counted per patient; a nasal swab is considered a screening swab)

• MRSA cases present on admission (MRSA known at the time of admission or MRSA isolated from screening swab or material collected within the first two days after admission)

• Nosocomial MRSA cases (MRSA isolated from swab or material collected from day three, i.e. later than day two, after admission)

• MRSA hospital days (assessment as defined above)

Based on the data submitted to the National Reference Center for the Surveillance of Nosocomial Infections (NRZ), the following measures are calculated:

• Overall prevalence (overall number of MRSA cases per 100 patients)

• Admission prevalence (number of MRSA cases present on admission per 100 patients)

• Incidence of nosocomial MRSA cases (number of nosocomial MRSA cases per 100 patients)

• Incidence density of nosocomial MRSA cases (number of nosocomial MRSA cases per 1000 patient days)

• Mean MRSA disease burden in the hospital (number of MRSA patient days per 100 patient days)

• Screening frequency (number of nasal swabs per 100 patient admissions; only one swab test during the hospital stay counts per patient)

These data are then analyzed at the German National Reference Center (NRZ); subsequently, the participating hospitals are provided with their own figures as well as the reference data.

For the period 2006–2021, we report the screening frequency in addition to these measures of MRSA epidemiology. For our data evaluation, we divided the screening frequency into four groups (categories/quartiles) of equal size, based on the frequency of nasal swabs performed. By analyzing overall MRSA prevalence and MRSA incidence density in the various categories, we were able to assess possible associations between epidemiological measures and screening frequency. Differences between categories were tested for significance using the Kruskal-Wallis test. For both overall MRSA prevalence and nosocomial MRSA incidence density, separate regression models were run to describe the temporal trend. Data analysis was performed using the open-source R statistical software (11).

We extrapolated the results of the MRSA surveillance data to all hospitals in Germany to estimate the overall burden of MRSA case in 2019 (the last calendar year before the start of the SARS-coronavirus 2 pandemic). German national hospital statistics data were used for this calculation (12)

Results

Since participation in KISS is voluntary, the number of reporting hospitals has fluctuated from year to year. Following the start of the MRSA module in 2006, 110 hospitals participated in this component. Since then, the number of participating hospitals has increased significantly, reaching a peak of 623 hospitals in 2018 which is about one third of all German hospitals. In 2021, 525 hospitals participated in the MRSA module.

Table 1 shows the development of MRSA epidemiology in German hospitals during the 16-year period from 2006 to 2021. In the pre-pandemic year 2019, the number of patient days included in the MRSA module corresponded to 41% of all patient days in German hospitals (the total number of patient days in 2019 was 139 million). On the assumption of an equal distribution of MRSA patient days in hospitals, the total number of MRSA patient days in Germany in 2019 was approximately 1.7 million, accounting for 1.2% of all annual patient days in German hospitals.

In 2012, overall MRSA prevalence peaked with 1.04 cases per 100 patients, remaining for several years on this level before starting to continuously and significantly decrease in 2017 to 0.72 (-32%) and 0.54 (-50%) in 2019 and 2021, respectively (p<0.001). Admission prevalence and mean daily MRSA disease burden also followed this trend. The incidence of nosocomial MRSA cases declined steadily from 0.26 per 100 patients in 2006 to 0.04 per 100 patients in 2021. Nosocomial incidence density decreased by an average of 12% per year and was 0.06/1000 patient days in 2021 (table 1). Figure 1 provides an overview of the development of MRSA epidemiology in hospitals and the frequency of screening (nasal swabs) for the period from 2006 to 2021.

Table 1

Parameters of participating hospitals and data on MRSA epidemiology in German hospitals, 2006 to 2021 (For epidemiological measures, such as prevalence, incidence, disease burden, and screening frequency, the respective mean values per year are shown [MRSA-KISS module])

Year	Number (n) of hospitals	Patient hospital days, total (n)	MRSA patient days (n)	MRSA cases (n)	Nosocomial MRSA cases (n)	MRSA cases present on admission (n)	Patients, total (n)	Prevalence (MRSA cases/ 100 pts)	Admission prevalence (MRSA cases poa/100 pts)	Nosocomial incidence (noso. MRSA cases/100 pts)	Nosocomial incidence density (noso. MRSA cases/1000 patient days [pds])	Mean MRSA disease burden (MRSA pds/total pds*100	Screening frequency (nasal swabs/100 pts)
2006	110	14 717 772	202917	13 171	3978	9193	1516083	0.87	0.61	0.26	0.27	1.38	5.78
2007	148	18161181	299 968	19 267	5 068	14 068	2197826	0.88	0.64	0.23	0.28	1.65	6.34
2008	191	23 724809	416 702	25 495	5801	19 694	3181412	0.80	0.62	0.18	0.24	1.76	8.17
2009	230	26 011 907	488 306	32 491	6 172	26 319	3399475	0.96	0.77	0.18	0.24	1.88	10.70
2010	303	32230693	633 078	44 411	7 151	37 260	4280846	1.04	0.87	0.17	0.22	1.96	14.60
2011	350	35790741	685 457	50 875	7 072	43 803	5218411	0.97	0.84	0.14	0.20	1.92	15.45
2012	400	40 917154	795 054	59 992	7617	52375	5750417	1.04	0.91	0.13	0.19	1.94	21.77
2013	459	46 048 614	873 582	67 641	7 416	60225	6622224	1.02	0.91	0.11	0.16	1.90	25.73
2014	492	48 981 503	882 793	70 891	7021	63870	7189343	0.99	0.89	0.10	0.14	1.80	30.00
2015	535	52 169 289	901 926	77 479	6 734	70745	8031532	0.96	0.88	0.08	0.13	1.73	33.45
2016	572	57 041 683	954 744	81 986	6 408	75578	8571016	0.96	0.88	0.07	0.11	1.67	38.76
2017	591	58314151	886016	79 869	5 909	73960	8884668	0.90	0.83	0.07	0.10	1.52	41.14
2018	623	58 780 627	827 669	76746	5 303	71443	9121328	0.84	0.78	0.06	0.09	1.41	43.05
2019	610	57 413703	683 536	64 832	4 221	60611	8964243	0.72	0.68	0.05	0.07	1.19	43.46
2020	583	48 397 647	496 998	48 216	3 017	45199	7593173	0.63	0.60	0.04	0.06	1.03	43.46
2021	525	44 882 632	386 778	37 715	2 503	35212	6967104	0.54	0.51	0.04	0.06	0.86	42.52

Open in a separate window

n, number; pts, patients; poa, present on admission; noso., nosocomial; pds, patient days

For the further analyses, the frequency of nasal swabs per 100 hospital admissions was used as a marker for screening frequency and divided into four equally sized categories (quartiles 1–4). Since the screening frequencies varied from year to year, the categories differed between years/time periods. From 2006–2018, a steady increase in screening frequencies was observed; in 2013 a nasal swab was obtained from at least one quarter of patients for the first time; from 2019, screening frequencies have plateaued on a high level (figure 1). The analyses included the year 2021 and the period 2013–2021. The categories used are presented in Table 2.

Open in a separate window

Figure 1

Annual MRSA prevalence on admission; nosocomial MRSA incidence and MRSA screening frequency (nasal swabs) in German hospitals in the period 2006–2021 (MRSA-KISS module)

MRSA, methicillin-resistant Staphylococcus aureus; pts, patients

In 2021, hospitals with high screening frequencies had a significantly higher MRSA admission prevalence compared to hospitals with a low screening frequency (Kruskal-Wallis test p<0.0001). However, the same observation was not made for nosocomial incidence density, as it remained stable across all screening categories (p = 0.8402) (table 3). When comparing these results with data from the 2013–2021 period, it can be seen that nosocomial incidence density was stable across all screening categories, consistent with the findings for the year 2021 alone (table 3).

Table 3

Distribution of MRSA rates by hospital screening frequency in the year 2021 and the period 2013–2021

Screening category (nasal swabs per 100 pts.)	Number of hospitals	Screening frequency (nasal swabs per 100 pts)	Admission prevalence (MRSA cases poa/100 pts)	Nosocomial incidence density (nosocomial MRSA cases/1000 patient days)
2021
≤ 24	119	14.33 (9.87–20.32)	0.28 (0.19–0.41)	0.04 (0.01–0.07)
24 to ≤ 40	102	32.36 (27.98–37.59)	0.38 (0.27–0.53)	0.04 (0.02–0.07)
40 to ≤ 64	112	48.89 (44.31–54.16)	0.53 (0.38–0.76)	0.04 (0.02–0.08)
> 64	115	85.67 (73.67–97.80)	0.61 (0.45–0.89)	0.04 (0.01–0.06)
Not stated	77	Not stated	0.40 (0.26–0.72)	0.03 (0.00–0.07)
All hospitals in 2021	525	40.36 (23.49–64.39)	0.43 (0.29–0.66)	0.04 (0.01–0.07)
2013–2021
≤ 19	1062	9.88 (4.92–14.35)	0.43 (0.27–0.71)	0.06 (0.02–0.11)
19 to ≤ 37	1081	28.23 (23.31–32.48)	0.60 (0.41–0.89)	0.06 (0.03–0.11)
37 to ≤ 58	1051	46.00 (41.26–51.34)	0.77 (0.53–1.14)	0.06 (0.03–0.12)
> 58	1071	81.70 (67.25–96.70)	0.93 (0.68–1.38)	0.07 (0.03–0.13)
Not stated	725	Not stated	0.66 (0.40–1.03)	0.07 (0.02–0.13)
All hospitals in 2013–2021	4990	36.76 (19.03–58.18)	0.68 (0.43–1.04)	0.06 (0.03–0.12)

Open in a separate window

The median and the interquartile range (lower and upper quartile) are shown. MRSA-KISS module, Germany; pts, patients

Discussion

A continuous decrease in nosocomial MRSA incidence is noted for the entire observation period and, since 2017, also a reduction in the overall and admission prevalence of MRSA (table 1). Therefore, the epidemiological trend observed in other surveillance systems for Germany and Europe can also be applied to the MRSA incidence in German hospitals (5, 6, 13). Similar to Germany, a statistically significant reduction in the incidence of invasive MRSA infections has been shown in the United States. From 2005–2013, the nosocomial incidence of bloodstream infections caused by MRSA strain USA 100 decreased from 6.1 to 0.9/100 000 population and for strain USA 300 from 1.5 to 0.6/100 000 population (14). Globally, MRSA rates in S. aureus isolates from blood or cerebrospinal fluid vary widely; in the period 2000–2016, for example, it ranged between 0% and 98% in the Asia-Pacific region (15). Figure 2 provides an overview of 2019–2020 MRSA rates for selected countries (6, 16, 17). Multiple pathogen-specific and host-specific factors as well as the heterogeneity of health care and surveillance systems worldwide have been identified as causes of variation in MRSA rates.

Significantly fewer patients were treated in German hospitals in 2022 and 2021 compared to pre-pandemic years (12). The MRSA-KISS data also show that 8% fewer patients received care in 2021 compared to 2019 (last pre-pandemic calendar year) (table 1). Elective hospital admissions, in particular, were deferred due to the SARS-CoV-2 pandemic. By contrast, patients with underlying chronic diseases who, for example, required dialysis or devices which are permanently left in place (such as an indwelling urinary catheter), continued to be admitted. Patients receiving dialysis on a regular basis or with indwelling urinary catheter or open wounds are at a higher risk of MRSA colonization (18). Nevertheless, admission prevalence continued to decline during the pandemic years. A meta-analysis concerned with the incidence of antimicrobial resistance in hospitals worldwide showed that the COVID-19 pandemic had no impact on the incidence density of MRSA (19); in the United States, the overall case rates for MRSA bloodstream infections remained stable between 2018 and 2020 (20). These observations made internationally during the years of the pandemic are reflected in the MRSA-KISS data.

Various causes have been attributed to the generally noted decline in MRSA. Apart from a potentially reduced biological fitness of drug-resistant S. aureus strains compared to drug-sensitive S. aureus, as well as a change in how antibiotics are used (21– 23), the decline in MRSA could be due to various bundles of infection prevention and control (IPC) measures. In 1999 and 2014, the Commission for Hospital Hygiene and Infection Prevention (KRINKO, Kommission für Krankenhaushygiene und Infektionsprävention) published recommendations for MRSA-related IPC measures to be taken in hospitals and care facilities (18, 24). These recommendations include enhancing basic hygiene and hand disinfection compliance. In addition, KRINKO supports targeted admission screening for patients at risk of MRSA and recommends the preparation of an infection control plan for each hospital, taking into account the local situation and country-specific recommendations for MRSA admission screening (18).

Notwithstanding this recommendation and a steady decline in MRSA incidence, screening frequency has increased in German hospitals over time since 2006 to remain at a high level since 2019 (figure 1). One reason for this could be that targeted screening of patients with an increased risk of MRSA—for example, re-admissions with known history of MRSA or admissions or transfers from facilities with known endemic or suspected MRSA occurrence (18, 24)—is considered to be more time-consuming to implement than screening entire departments or the entire hospital. What is more, MRSA detected during hospitalization is classified as nosocomially acquired in patients who have not undergone admission screening. To address this issue, admission screening has been expanded in some hospitals. Furthermore, hospitals could use a comprehensive MRSA screening strategy for promotional purposes, as the media keeps reporting on alleged successes of universal admission screening. In these reports, however, reference is often made to the Netherlands, despite the fact that only risk-based screening is performed there (25).

Finally, it appears that there is still a widespread perception that universal admission screening is an evidence-based IPC measure. However, this view is refuted or critically discussed in various studies (26– 30). Likewise, MRSA-KISS figures reported in recent years indicate that hospitals with low or moderate screening frequencies did not see an increase in the number of nosocomial MRSA cases detected. For example, for the year 2021, the admission prevalence for a hospital was found to be significantly higher if the hospital was in the 4^th quartile with respect to the frequency of admission swabs instead of the first quartile. However, the incidence density of nosocomial MRSA was found to be stable at 0.04/1000 patient days, independent of the screening frequency (table 3). Thus, the question arises whether a high screening frequency is necessary at all. Rather, IPC measures appear to be increasingly implemented successfully to prevent nosocomially acquired MRSA cases. Risk-adapted targeted MRSA admission screening can help to conserve staff capacities and laboratory resources. Therefore, such a screening strategy appears to be useful, especially in view of the stable epidemiological situation.

Conclusion

A significant decrease in MRSA cases in German hospitals has been observed for the period 2006–2021. The incidence density of nosocomial MRSA is not associated with screening frequency. This means that the probability of nosocomial transmission of MRSA is not increased in hospitals with a low number of nasal swabs per 100 patients. Given the lack of a preventive effect and the resource consumption associated with a high screening frequency, a targeted screening strategy focused on patients at increased risk of MRSA should be preferred over universal MRSA admission screening.

Acknowledgments

Footnotes

References