Hand hygiene opportunities and consumption of alcohol-based handrub across nursing homes of canton of Vaud, Switzerland Authors: Emmanouil Glampedakis, Coralie Riccio, Firmino Battistella, Patricia Cuiña Iglesias, Francisco-Javier Garcia Sanchez, Marie-Catherine Snoussi, Béatrix Sobgoui, Marie Immaculée Nahimana Tessemo Published: 2026 ABSTRACT BackgroundOptimal hand hygiene (HH) is necessary to prevent healthcare-associated infections among nursing home (NH) residents. We aimed to quantify the number of HH opportunities per resident-day in NHs and assess institutions’ alcohol-based hand rub (ABHR) consumption based on the HH needs of their population.MethodsCross-sectional study between 2022 and 2024 with the participation of 117 NHs (95%) and a random sample of 1633 residents. For each resident we accessed the daily care schedule, attributed 2–4 HH opportunities per nursing activity and collected the care dependency score. We estimated the average number of HH opportunities per resident-day across NHs using linear mixed-effects modeling. We also compared theoretical and observed ABHR consumption on an institutional level.ResultsThe average amount of HH opportunities was 21.65 per resident-day and the need was highly dependent on the care dependency score (β = 2.93; 95% confidence interval: 2.77–3.10, p < 0.01). Median ABHR was 9.43 L per 1000 resident-days. Overall, across all institutions ABHR consumption was 24% of the expected as most institutions consumed up to 25% of the expected amount of ABHR.ConclusionsIn this setting, HH opportunities are higher than previously assumed and significantly associated with care dependency. Most NHs perform poorly and should be prioritized in HH improvement actions like education and auditing with direct feedback.Supplementary InformationThe online version contains supplementary material available at 10.1186/s13756-026-01776-x. Introduction Nursing home (NH) residents are at increased risk of healthcare-associated infections (HAI) because of prolonged residence in closed environments and overall frailty [1]. HAIs have significant implications for residents, including increased mortality, diminished quality of life, and the need for transfer to acute care settings for treatment [2]. Recent post-pandemic evidence further indicates that mortality attributable to infectious diseases remains high among NH residents [3]. In Switzerland, a recent point prevalence survey conducted outside of the respiratory season reported a HAI prevalence of 2.2% [4], comparable to the European average [5]. Although this rate is lower than that observed in hospitals, it is expected to rise substantially during the winter months, when respiratory and gastrointestinal outbreaks prevail [6]. Transmissions are facilitated by close living conditions and through contact with healthcare workers’ (HCWs) contaminated hands during nursing care. Although hand hygiene (HH) using alcohol-based handrub (ABHR) is an effective measure for preventing HAIs and represents a cornerstone of infection prevention bundles in NHs [7], adherence remains suboptimal [8]. Most institutions have HH protocols [4, 5], yet compliance varies widely and overall falls far below the 80% target recommended by WHO [8, 9]. Low HH adherence increases the risk of large-scale HAI outbreaks [10] and facilitates the transmission of multidrug-resistant organisms in NHs [11]. Many current studies assess HH compliance in NHs through direct observation of HCWs [7]. Simpler, less time-consuming approaches use surrogate indicators that indirectly reflect HH compliance without requiring direct observation. Some reports analyze institutional-level ABHR consumption [5], while others use ABHR consumption to estimate daily HH opportunities per resident-day [12]. However, the former are difficult to interpret without data on the actual HH needs per resident-day, and the latter do not provide information on those needs. Very few studies [12] have reported the actual daily number of HH acts required for the care of NH residents. Such data could serve as a basis for estimating more accurate ABHR requirements in NHs, which can then be compared with actual consumption over time to indirectly assess HH performance in institutions. The aim of this study was to quantify the daily number of hand decontaminations required for the nursing care of NH residents in our region. Moreover, we aimed to indirectly assess HH performance in participating NHs by comparing estimated needs with actual ABHR consumption. Methods Setting We conducted a cross-sectional study across NHs in the canton of Vaud, Switzerland between 2022 and 2024. All nursing homes of Vaud providing on-site medical care for persons aged 65 years and older were invited to participate. The canton of Vaud is one of Switzerland’s largest cantons, comprising 123 NHs and approximately 6782 residents in 2023 (68% female; mean age 87.7 years). The NHs in Vaud operate as either private non-profit or private for-profit facilities. They are categorized into three main types: geriatric homes (for residents with age-related physical dependency), psychogeriatric homes (for residents with neurodegenerative psychiatric conditions), and mixed facilities. They are further categorized by location as urban (within towns or cities), intermediate (on their outskirts), or rural. Residents reside in single (80%) or double-bed (20%) rooms [13] and have access to shared spaces like a common living and dining rooms. Care is offered by nurses, nursing assistants and trainees of these professions. Infection prevention and control (IPC) recommendations and activities are coordinated by the cantonal infection prevention and control unit of Vaud (HPCi Vaud). HPCi Vaud, which is integrated within the cantonal doctor’s office, promotes best practices, provides education on IPC, monitors IPC indicators, and conducts audits to ensure safe care for residents and the protection of HCWs. HCWs have open access to IPC protocols and recommendations including HH in a cantonal online platform (www.hpci.ch). Sample In each participating NH, a random sample of 10–20 residents, depending on the facility’s size, was selected (Supplementary Figure S1). Stratified random sampling was used to ensure appropriate representation of populational care dependency proportions. A member of HPCi Vaud, together with the local nursing team, reviewed the care schedules of residents in the sample and recorded their care dependency scores. Data collection The care dependency score, available for every participating resident, was an ordinal variable ranging from 1 to 12, with higher values indicating greater care needs [14]. A score of 1 corresponded to 0–20 min of daily nursing care, with each additional point representing a further 20 min of care. Care schedules listing all daily nursing care activities, whether documented in electronic health records or on paper, were available for each participating resident and were reviewed with the local team to ensure their accuracy and relevance on the day of the assessment. For each activity listed in the care schedule, we systematically assumed either two or four HH opportunities, depending on the type of nursing task (Appendix, HH opportunities collection tool). For individual care activities, we assumed two HH opportunities, corresponding to one action before resident contact or a clean/aseptic procedure and one after resident or body fluid or resident’s environment contact, in line with the WHO’s five moments framework for HH. When several care activities were grouped within a single care episode, we deliberately estimated four opportunities to better reflect the increased number of potential HH indications. The total number of daily HH opportunities was then estimated for each resident in the sample. At the institutional level, data were collected on the number of beds, number of residents, facility type, and location. In addition, NHs reported the volume of ABHR consumed, expressed in liters, over a 90-day period following the study. These quantities also included ABHR volumes from healthcare workers’ portable bottles. The amount consumed was calculated by subtracting the stock at the end of the surveillance period from the sum of the stock available at the beginning of the period and the amount purchased during the same timeframe. Statistical and other analyses Descriptive analyses included numbers, percentages, medians, interquartile ranges (IQR), means and standard deviations (SD), as appropriate. Linear mixed-effects models were applied to the sample to estimate the average number of HH opportunities across NHs in our region. We further examined the association between the care dependency score and the number of HH opportunities per resident-day in the sample of residents, adjusting for nursing home size, type and localization. These models account for the clustering of observations within facilities (NHs as random intercepts). As only the care dependency score was significantly associated with the number of HH opportunities, and population scores were available in three-unit bands (1–3, 4–6, 7–9, 10–12), a separate univariable model with 3-unit increment dependency score as predictor and HH acts per resident-day as outcome was fitted on the sample (NHs as random intercepts). This model, taking into account the care dependency category (1–3, 4–6, 7–9 or 10–12) and the institution, was used to estimate the number of HH opportunities per resident per day in the whole population of each NH. Subsequently, the theoretical ABHR volume per resident per day was calculated assuming 3 mL per HH act. The volume of 3 mL per hand hygiene act was selected because it aligns with previous studies and surveillance protocols from Europe [12, 15], thereby enhancing comparability across settings. These volumes were aggregated to derive the theoretical daily requirement per facility, then multiplied by 90 days to estimate the expected consumption over the observation period. Theoretical and observed ABHR consumption over 90 days were compared, and institutions were classified by performance as follows: A + (> 100%), A (75–100%), B (50–75%), C (25–50%), and D (0–25% of the theoretical requirement). Finally, NHs were also categorized by performance using a fixed number of HH opportunities (5 per resident-day) as previously proposed [15]. Results were considered statistically significant at p ≤ 0.05. All analyses were conducted using R statistical software (version 4.3.3; R Foundation for Statistical Computing, Vienna, Austria) within the RStudio environment. Ethical statement No personal data from residents in nursing homes were used in this project. The study was conducted in line with the local ethics committee’s criteria for quality improvement projects; therefore, ethics approval was not required. Results Sample characteristics One hundred and seventeen NHs (95%) were included in the study, most of which were urban and geriatric facilities (Table 1). Forty-five NHs had up to 40 beds, 54 had between 40 and 79 beds, and 18 had 80 beds or more. Among the 5865 residents present in these institutions on the day of the study, 1633 (28%) were included. The mean care dependency score was 9.05 (SD: 2.54), while half residents had a score of 10 or more (IQR: 7–11), corresponding to a median of 180–200 min of daily nursing activities. Care dependency score distributions were similar between residents included in the study and those not included (Supplementary Table 1).Table 1Characteristics of included nursing homesNursing home characteristics (N = 117)N (%)aFacility typeGeriatric68 (58)Psychogeriatric23 (20)Mixed26 (22)LocalizationUrban70 (60)Rural20 (17)Intermediate27 (23)Number of beds, median (IQR)48 (32–68)Occupied beds, median (IQR)100% (97–100%)Number of residents, median (IQR)47 (32–66)a: if not stated otherwiseN, Number; IQR, Interquartile range Table 1Characteristics of included nursing homesNursing home characteristics (N = 117)N (%)aFacility typeGeriatric68 (58)Psychogeriatric23 (20)Mixed26 (22)LocalizationUrban70 (60)Rural20 (17)Intermediate27 (23)Number of beds, median (IQR)48 (32–68)Occupied beds, median (IQR)100% (97–100%)Number of residents, median (IQR)47 (32–66)a: if not stated otherwiseN, Number; IQR, Interquartile range Characteristics of included nursing homes a: if not stated otherwise N, Number; IQR, Interquartile range Hand hygiene opportunities per resident-day The mean daily number of HH opportunities in the sample was 21.6 (median 20, IQR: 13–28). Of all recorded HH opportunities, 43% were associated with morning and evening care packs. Outside of these, 39% were related to medication administration, 21% to hygiene care, 16% to mobilization assistance, 12% to technical care, 11% to comfort care, and 1% to housekeeping activities. Multilevel modeling (null model) predicted 21.65 HH opportunities per resident-day on average across NHs (Table 2). The distributions of HH opportunities per resident-day according to the care dependency score are shown in Fig. 1. The regression analysis showed that care dependency score was significantly associated with HH opportunities with one-unit increase in the score followed by an increase of 2.93 daily HH opportunities on average (95% confidence interval (CI): 2.77—3.10, p < 0.01). Adjustment for factors including NH size, type and localization did not change this relationship with the outcome as shown in Table 2.Table 2Univariable and multivariable models on factors associated with daily hand hygiene opportunitiesPredictorObserved HH opportunities, Median (IQR)β (95% CI)p—valueβ (95% CI)p—valueUnivariable modelsMultivariable model(Null model)20 (13–28)21.65 (20.69–22.62) < 0.01Care dependency score2.93 (2.77–3.10) < 0.012.93 (2.77–3.10) < 0.01Below median14 (10–20)Above median26 (20–34)Facility typeGeriatric*20 (12–28)––––Psychogeriatric24 (14–32)2.92 (0.42–5.41)0.020.82 (− 1.42–3.07)0.47Mixed20 (12–28)0.09 (− 2.23–2.42)0.940.41 (− 1.83–2.64)0.72LocalizationUrban*20 (13–28)––––Rural22 (12–30)0.79 (− 1.92–3.50)0.570.03 (− 2.46–2.52)0.98Intermediate20 (14–26) − 1.20 (− 3.55–1.15)0.32 − 1.80 (− 3.90–0.31)0.1Number of beds − 0.02 (− 0.05–0.02)0.310.002 (− 0.03–0.04)0.89Below median20 (12–28)Above median21 (14–28)*Taken as reference in the models, CI, Confidence intervalFig. 1Distribution of daily numbers of hand hygiene opportunities according to the care dependency score Table 2Univariable and multivariable models on factors associated with daily hand hygiene opportunitiesPredictorObserved HH opportunities, Median (IQR)β (95% CI)p—valueβ (95% CI)p—valueUnivariable modelsMultivariable model(Null model)20 (13–28)21.65 (20.69–22.62) < 0.01Care dependency score2.93 (2.77–3.10) < 0.012.93 (2.77–3.10) < 0.01Below median14 (10–20)Above median26 (20–34)Facility typeGeriatric*20 (12–28)––––Psychogeriatric24 (14–32)2.92 (0.42–5.41)0.020.82 (− 1.42–3.07)0.47Mixed20 (12–28)0.09 (− 2.23–2.42)0.940.41 (− 1.83–2.64)0.72LocalizationUrban*20 (13–28)––––Rural22 (12–30)0.79 (− 1.92–3.50)0.570.03 (− 2.46–2.52)0.98Intermediate20 (14–26) − 1.20 (− 3.55–1.15)0.32 − 1.80 (− 3.90–0.31)0.1Number of beds − 0.02 (− 0.05–0.02)0.310.002 (− 0.03–0.04)0.89Below median20 (12–28)Above median21 (14–28)*Taken as reference in the models, CI, Confidence interval Univariable and multivariable models on factors associated with daily hand hygiene opportunities *Taken as reference in the models, CI, Confidence interval Distribution of daily numbers of hand hygiene opportunities according to the care dependency score Alcohol-based handrub consumption One hundred and sixteen NHs (94%) provided ABHR consumption data. Over 90 days, the model predicted a cumulative 33,454 L, whereas observed consumption was 7931 L (24% of expected). Predicted consumption showed a median 62.3 L (IQR 57–68.4) per 1000 resident-days, versus an observed median 9.43 L (6.58–16.4) per 1′000 resident-days. Observed consumption was slightly higher in psychogeriatric institutions (median 10.2 L; IQR 7.31–25) than in geriatric (9.43; 6.55–15.5) and mixed-type (8.13; 6.67–15) facilities (p = 0.60). Using the model-based benchmarks, most institutions (84, 72%) fell into the lowest performance category D (Fig. 2). In contrast, when applying a fixed number of 5 HH opportunities per resident-day the A + category collected the highest number of institutions (32, 28%; Fig. 2).Fig. 2Categorization of nursing homes by alcohol-based handrub consumption performance. Letters on the base of the bars correspond to performance categories. A + (> 100%): 3 vs 32, A (75–100%): 2 vs 15, B (50–75%): 4 vs 28, C (25–50%): 23 vs 27, D (0–25%): 84 vs 14 institutions. HH: Hand hygiene acts, NH: Nursing home Categorization of nursing homes by alcohol-based handrub consumption performance. Letters on the base of the bars correspond to performance categories. A + (> 100%): 3 vs 32, A (75–100%): 2 vs 15, B (50–75%): 4 vs 28, C (25–50%): 23 vs 27, D (0–25%): 84 vs 14 institutions. HH: Hand hygiene acts, NH: Nursing home Discussion We aimed to quantify the daily HH opportunities required for the care of NH residents in our region. The mean was 21.65 HH opportunities per resident-day, with values increasing markedly with care dependency. ABHR consumption was below expected needs, indicating substantial room for improvement across regional institutions. Currently, most reports either rely on direct HH observations, which are time-consuming and difficult to scale, or on ABHR consumption without analyzing the underlying HH needs, making interpretation challenging. Recent reports have estimated performed HH acts per resident-day from ABHR consumption, assuming 3 mL per act. Their findings are concerning as most institutions appear to perform fewer than two HH acts per resident-day [12]. Given that a single care episode typically entails at least two opportunities (before and after resident contact), this level is likely insufficient. While these reports are valuable in highlighting shortfalls in HH performance, they add little to clarifying the actual daily HH needs and therefore do not substantially narrow the existing knowledge gap. In contrast, our HH needs-based approach suggests that required HH opportunities may be substantially higher than commonly assumed, with an estimated mean of 21.65 per resident-day. Previous studies have reported an average of approximately 2.5 HCW–resident interactions or care activities per resident per hour in long-term care units [16, 17], which, assuming at least two HH opportunities per interaction, supports the notion that the overall number of HH opportunities in these settings may be higher than previously estimated. Our results should be interpreted considering the high care dependency in our population (~ 3 h of nursing per resident-day) that could also explain the discrepancy with prior estimates [12]. Moreover, some surveillance methodologies have assumed, probably for simplicity and ease of implementation, a fixed need of 4–5 HH acts per resident-day in NHs and other long-term care facilities and use this to derive theoretical ABHR volumes for comparison with consumption data [15]. Beyond the substantially higher requirement observed here, our results suggest that a fixed number is inappropriate, as HH opportunities vary with care dependency. Both univariable and multivariable models showed that each one-unit increase in care dependency score was associated with an average increase of 2.93 HH opportunities per resident-day. These results better reflect the reality and heterogeneity of care dependency, and the consequent HH needs, in our region’s resident population. Because resident profiles may vary across settings, absolute estimates should be interpreted cautiously; however, we propose a scalable HH needs-based framework that can be tested and refined across different institutional, geographical and resource settings prior to broader implementation. For instance, while the care dependency score used in this study is specific to our region, it can be easily translated in nursing care time, so that the proposed framework could be adapted to other healthcare systems using locally available measures of resident care needs. Future prospective observational studies combining resident care activity measurements, HH opportunities, ABHR consumption, and HH compliance levels would be valuable to further validate and calibrate this approach prior to broader implementation. Although HH protocols are among the most widely available patient-safety recommendations [4, 5], compliance in nursing homes remains low [8, 9]. Consistent with this, observed ABHR consumption represented only 24% of the modelled requirement for this population’s care needs. The performance categorization indicated that most facilities fell into the lowest category (D; Fig. 2). By contrast, when applying a fixed number of HH acts per resident-day, many institutions (41%) were classified in the highest performance categories, with several appearing to consume more ABHR than necessary (A + , A; Fig. 2), an implausible pattern that complements the above-mentioned limitations of low fixed HH assumptions per resident-day. Overall, despite poor performance and clear scope for improvement, observed ABHR consumption in this cohort exceeded the latest European median (9.43 vs 4.9 L per 1000 resident-days)[5]. Interestingly, ABHR consumption in our region was closer to the value reported in Germany (11.4 L/1000 resident-days) [5]. However, comparability across geographical and resource settings remains limited due to differences in resident profiles and care needs, highlighting the need for more in-depth studies to accurately interpret surrogate markers such as ABHR consumption. The principal strength of this study is that it proposes a needs-based methodology for estimating HH requirements and interpreting ABHR consumption data. Drawing a representative sample of residents across a large number of institutions in our region, documenting care schedules alongside care dependency scores, we estimate HH opportunities for each score category, translate these into ABHR volumes, aggregate results at facility levels, and compare estimated needs with observed consumption over a period of time to support better interpretation and benchmarking. While more complex to implement than simple HH surrogate indicators such as usual ABHR consumption, our framework provides a more nuanced estimation of ABHR needs taking into account resident profiles. Moreover, it might also render surrogate indicators more tangible and interpretable for institutions and decision-makers. Its impact on feasibility, acceptability, and HCW behaviors with respect to HH should be explored in future studies. Our study has several limitations. Firstly, due to its cross-sectional design, the findings are based on the care schedules of residents available on the day of the study, which may be subject to change. Nonetheless, we are confident that the care dependency scores used were up to date, as they are linked to institutions’ reimbursement claims. Secondly, we deliberately attributed two HH opportunities to most care activities, and four where multiple care acts are grouped together, such as morning and evening care packs; these numbers may differ under real-world conditions. Although the estimated number of HH opportunities depends on predefined assumptions (such as the number of opportunities assigned per care activity individually or when grouped together), the main objective of this approach is to provide a standardized proxy of HH needs rather than an exact absolute count. Importantly, variations in these parameters would proportionally affect absolute estimates but are unlikely to modify the observed association between care dependency and HH needs or the relative comparison between institutions. Moreover, the study did not include resident-level demographic, clinical and other characteristics that could influence HCW–resident interactions and, consequently, HH practices, such as isolation status [17]. These factors would be valuable to explore in future research. However, in our setting, isolation measures are exceptional, and most residents are managed under standard precautions; therefore, their impact on our findings is expected to be minimal. As ABHR consumption was used as a proxy for HH performance, the study lacks information on the specific HH moments associated with poor adherence. For example, prior data indicate lower adherence before contact with the resident [8, 9, 18] and before clean or aseptic procedures [8], with somewhat better adherence after resident contact [8, 18]. Such granular insights could help prioritize training activities for specific groups of HCWs and should be encouraged in future research. Moreover, we assumed a fixed volume of 3 mL per HH act, to increase comparability with previous reports [12], although there is no uniformly accepted volume of ABHR to ensure complete coverage of all sizes of hands [19] and the actual quantity used in practice may be lower [20]. In line with this, several studies suggest that HCWs often prefer volumes below our assumed 3 ml [21, 22] which very likely have inflated theoretical ABHR consumption, therefore influencing the theoretical versus real ABHR consumption comparisons in our study. Future research could consider using HH volume estimates that reflect locally preferred practices to better represent institutional ABHR use. Furthermore, ABHR consumption was reported by NHs, which could have affected measurement and reporting accuracy. Another limitation was the absence of actual resident-days for calculations, as such data were unavailable from institutions. We therefore assumed that the population surveyed on the study day remained unchanged for 90 days for estimation purposes. Although imperfect, this assumption is plausible given the long stays of most residents and the high bed occupancy observed on the study day; however, it may not hold for institutions with high resident turnover. Our study did not include data on alternative HH practices, such as handwashing with soap and water, nor on factors such as inappropriate glove overuse, product wastage, temporary dispenser unavailability, or limited ABHR accessibility at the point of care, all of which may contribute to low observed ABHR consumption. Finally, we lacked data on the proportion of healthcare workers using portable ABHR bottles. We therefore hypothesize that institutions with low availability of such dispensers may exhibit lower ABHR consumption, as accessibility to handrub is a key determinant of hand hygiene performance [23]. Conclusion In conclusion, this study contributes to the existing methodological framework for the interpretation of surrogate indicators by proposing a pragmatic approach to quantifying HH needs per resident-day in NHs. In our region, an estimated 21.65 HH acts per resident-day were required on average, with demand strongly driven by residents’ dependency on care. Furthermore, the proposed method for estimating HH opportunities may be adopted to support monitoring and improvement of HH compliance in other long-term care settings. Our findings will be used to support the implementation of targeted HH compliance audits and training sessions in institutions of our region with suboptimal performance, with the aim of improving practices and enhancing resident safety in this highly vulnerable population. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2