Use of single-bed rooms may decrease the incidence of hospital-acquired infections in geriatric patients: A retrospective cohort study in Central Denmark region

Abstract

Objective

Patients accommodated in single-bed rooms may have a reduced risk of hospital-acquired infections (HAIs) compared to those in multi-bed rooms. This study aimed to examine the effect of single-bed accommodation on HAIs in older patients admitted to a geriatric ward.

Methods

A retrospective cohort study of patients admitted to geriatric wards in a university hospital in Central Denmark Region linked to a move to a newly built hospital, involving all consecutively admitted patients aged 65 years and over from 15 September to 19 December 2016 and a similar cohort admitted in the same three months in 2017. We compared the incidence of HAIs in patients in single-bed accommodation to those in multi-bed accommodation using retrospective review of electronic patient records, with all infections verified microbiologically or by X-ray with onset between 48 hours after admission to 48 hours after discharge from hospital.

Results

In total 446 patients were included. The incidence of HAIs in multi-bed accommodation was 30% compared to 20% in single-bed accommodation. The hazard ratio was 0.62 (95% Confidence Interval 0.43–0.91, p = 0.01) for single-bed accommodation. This finding remained robust after adjustment for age, sex, infection at admission, risk of sepsis, use of catheter, treatment with prednisone or methotrexate, and comorbidity index.

Conclusion

Accommodation in single-bed rooms appeared to reduce HAIs compared to multi-bed rooms in two geriatric wards. This finding should be considered as hypothesis-generating and be examined further using an experimental design.

Keywords

Hospital-acquired infection geriatrics single-bed rooms

Introduction

Nosocomial infections pose a serious risk to patient safety during hospitalization. With a prevalence of 5.7% in Europe, hospital-acquired infections (HAIs) account for an estimated 37,000 deaths and 16 million extra bed-days.¹ The most prevalent HAIs are urinary tract infection, surgical site infection, hospital-acquired pneumonia and bloodstream infection.² It has been estimated that 50–70% of the most frequent types of HAIs are reasonably preventable with current evidence-based strategies.³ These strategies rely on knowledge of the interactions of pathogens, patients, clinical practice, and the place where health care is provided.⁴

Hospital and ward design as a measure to prevent infection is widely discussed and several guidelines on hospital architecture and ward design have emerged. For instance, the Department of Health in England requires that any newly built hospital should have a minimum of 50% single-bed rooms. The policy is based, in part, on the assumption that single-bed accommodation lowers the risk of acquiring HAIs,⁵ as shown in a 2016 meta-analysis, which concluded that use of single-bed rooms decreased the overall risk of health care-associated colonisation and infection at a risk ratio of 0.55 compared to multi-bed rooms.⁶ However, other work found the overall evidence supporting single-bed accommodation in low-acuity settings lacking.⁷ A relatively larger number of higher quality studies were carried out in intensive care units (ICUs) and acute wards with only few in a general hospital setting.^8–18 Two studies outside ICUs showed positive results regarding the preventive effect of single-bed accommodation on HAIs.^19,20 These findings may indicate that the preventive effect from accommodating patients in single-bed rooms vary depending on the group of patients and type of wards.

Geriatric patients are highly susceptible to infections due to deterioration in immunity, various comorbidities, and an increased risk of hospitalization.²¹ The risk of a HAI increases linearly with age, with a prevalence of 11% in patients aged 75–85 years.²² The most prevalent types of HAIs in the older patients are respiratory infections and urinary tract infections and together they are associated with an overall mortality rate of 4%.²² Because of the higher risk of HAIs among geriatric patients it is plausible to assume that this group would benefit from accommodation in single-bed rooms. We only identified a single retrospective study which did not find a reduction in the number of HAIs among patients aged 65 or older accommodated in single-bed rooms compared to those in multi-bed rooms.¹² However, that study only included 166 patients with nine occurrences of HAIs. The present study aimed to examine if accommodation in single-bed rooms reduces the number of HAIs compared to multi-bed rooms in older patients admitted to geriatric wards.

Methods

Design, population and setting

The study was conducted as a retrospective cohort study conducted in one university hospital in Central Denmark Region. We took advantage of the move of the geriatric ward to a newly built hospital on 20 March 2017. The study population consisted of all patients admitted to either the old hospital with multi-bed rooms during 15 September to 19 December 2016, or the new hospital with single-bed rooms in the same period in 2017. The old geriatric wards consisted primarily of multi-bed rooms (5 three bed, 11 two bed, and 2 single bed; all with shared bath/toilets), whereas the new ward comprised of 32 single-bed rooms, each with their own bath/toilet. All rooms were equipped with sinks and hand sanitizer dispensers for staff use close to room entrance in both hospitals.

There were no changes in the hospital’s catchment area, admission criteria, or staff composition over the entire observation period. Terminally ill patients and those admitted more than once to the same hospital were excluded.

Data collection

Patients were identified and relevant data extracted electronically from the Central Denmark Region’s database for hospitalised patients. Data not available in the database were obtained from reviewing individual electronic patient records by two of the authors, a medical doctor and a clinical nurse specialist. Baseline information regarding age, sex, housing situation (own home or nursing home), body mass index (BMI), activities of daily living (ADL) as measured by the Barthel-100 Index, main hospitalization diagnosis, pulse, temperature, and respiratory rate were obtained upon admission. A Charlson Comorbidity Index (CCI) was calculated for every patient based on diagnosis recorded in their medical records. Two or more positive criteria of the Systemic Inflammatory Response Syndrome (SIRS) (temperature > 38 °C or <36 °C, pulse > 90bmp, respiratory rate > 20 breaths per minute, white cell count > 12×10⁹/L or <4×10⁹/L) and a clinical suspicion of an infection were taken as indicative of sepsis. Blood samples were drawn to measure haemoglobin, white blood cell count and C-reactive protein levels.

All diagnoses of infection were recorded. To qualify as a hospital-acquired infection, it had to be either microbiologically verified or, in the case of pneumonia, verified by X-ray. In a small number of cases, a significantly high level of C-reactive protein, combined with fever and clinical symptoms of infection that led to treatment with antibiotics was considered as sufficient to qualify as HAI. Additionally, the onset of infection had to occur between 48 hours after hospital admission and 48 hours after discharge from hospital. Tracking infection after discharge was possible due to a system of continued geriatric care outside of hospital, involving an early follow-up home visit by a geriatric team consisting of a physician and a nurse.²³ The team monitored and treated infections in the patients' home and recorded any examination, infection and treatment in the electronic patient record.

Statistical analysis

Based on a pilot study in our department, a group of 30 patients admitted in 2016 was compared with the first 30 patients admitted to the new wards. The prevalence of hospital-acquired infections was 20% in the old wards and 10% in the new wards. This meant that, at a power of 80% and level of significance at 5%, we would require 219 patients in each cohort to detect a statistically significant difference in HAIs between multi-bed and single-bed rooms.

To compare baseline data, we used Student’s t test or Wilcoxon’s rank sum test for continuous variables and Pearson’s chi-squared test or Fisher’s exact test for categorical variables. We counted the HAIs for each cohort. The counts were divided by the total number of hospital days for that hospital and multiplied by 1,000. A Cox regression model was used to compare the incidence of HAIs, adjusted for the following possible confounders measured at admission: age, sex, infection at admission, treatment with prednisone or methotrexate, use of catheter, and a score for risk of sepsis. Time at risk was considered from hospital admission to 48 hours after discharge. Where multiple HAIs arose in a single patient, only the first event was included in the analysis. Also, patients could be included twice; at first to the old hospital and then to the new hospital. Therefore, the Cox regression analysis was adjusted for this clustering. Robust variance estimates were calculated. A test of the proportional hazards assumption was supplemented with inspection of ’log-log‘plots. All statistical analyses used Stata software, version 15.1 (StataCorp LLC, College Station, Texas). P-values <0.05 were considered statistically significant.

Ethics review

The study was conducted according to the principles of the Declaration of Helsinki. Since the data were collected respectively using the patients' electronic health records, approval by The Central Denmark Region Ethical Committee was not required. The head of the geriatric department approved the study as a quality control project.

Results

In total, 446 patients were included in the study, involving 220 patients in multi-bed rooms in the old ward and 226 patients in single-bed rooms in the new ward. Table 1 present the baseline characteristics of the two cohorts. Patients in the old and new wards were statistically similar on all variables except for age (85.1 vs. 83.2, p = 0.01), sex (36% vs 46% males, p = 0.04), and the number of patients with infection at admission (42% vs. 54%, p = 0.01) presenting with two or more positive SIRS criteria at admission (16% vs. 24%, p = 0.02).

Table 1.

Baseline characteristics of 446 geriatric patients.

Baseline characteristic	Multi-bed roomsN = 220	Single-bed roomsN = 226	p value
Age, y (sd)	85.1 (6.9)	83.2 (7.9)	0.01
Gender, male (%)	79 (36)	103 (46)	0.04
Length of hospital stay, days (IQR^a)	9 (6 – 12)	9 (6 – 11)	0.50
Housing situation (%)
Own home	206 (94)	201 (89)
Nursing home	14 (6)	25 (11)	0.07
Charlson Comorbidity Index (%)
Low (score 0)	65 (29)	70 (31)
Moderate (score 1–2)	105 (48)	106 (47)
High (score >2)	50 (23)	50 (23)	0.93
Activities of daily living (MBI^b) (%)
Independent	27 (12)	21 (9)
Moderate dependent	75 (34)	73 (32)
Substantially dependent	62 (28)	65 (29)
Completely dependent	56 (26)	67 (30)	0.60
Infection at admission, total (%)	93 (42)	123 (54)	0.01
Urinary tract	36 (16)	46 (20)	0.28
Pneumonia	38 (17)	53 (23)	0.11
Gastritis (Clostridium difficile)	3 (1)	2 (1)	0.63
Sepsis	6 (3)	13 (6)	0.11
Other infections	10 (5)	9 (4)	0.77
SIRS^c ≥2 (%)	34 (16)	54 (24)	0.02
Ongoing treatment at admission (%)
Prednisolone	35 (16)	39 (17)	0.70
Methotrexate	5 (2)	5 (2)	0.97
Use of urinary catheter during hospitalization (%)	173 (79)	174 (77)	0.68
BMI^d (IQR)	24.7 (21.2 – 27.4)	24.1 (21.2 – 26.6)	0.27
Blood samples (IQR)
C-Reactive protein,^e mg/L	52 (16 – 99)	61 (20 – 125)	0.11
White blood cells,^e 10⁹/L	9.0 (6.7 – 11.8)	8.4 (4.3 – 11.4)	0.11
Hemoglobin,^f mmol/L	7.3 (6.6 – 8.2)	7.4 (6.8 – 8.3)	0.19

^aIQR: interquartile range with 25% and 75% percentiles.

^bModified Barthel Index-100.

^cSystemic Inflammatory response syndrome.

^dBody mass index.

^eC-Reactive protein and white blood cells measured within the period: 4 days before to 2 days after admission.

^fHemoglobin measured in the period: 4 days before to 3 days after admission.

A total of 141 HAIs were recorded in 110 patients during hospitalization and up to 48 hours after discharge. In 31 patients more than one HAI occurred. Table 2 presents the distribution of HAIs among the patient cohorts. Of those accommodated in multi-bed rooms, 66 patients acquired an infection (30%) compared to 44 patients in single-bed rooms (20%). The infection rate was 30.5 per 1,000 hospital days in multi-bed accommodation compared to 21.2/1000 in single-bed accommodation. The higher rate of HAI in multi-bed accommodation was driven, largely, by urinary tract infections.

Table 2.

Distribution of hospital-acquired infections in 446 geriatric patients in multi-bed rooms or single-bed rooms.

	Multi-bed roomsN = 220	Single-bed rooms N = 226	p value
Urinary tract infection (%)	41 (18.6)	21 (9.3)	0.004
Pneumonia (%)	25 (11.4)	28 (12.4)	0.74
Gastritis (Clostridium difficile) (%)	3 (1.4)	5 (2.2)	0.50
Sepsis	5 (2.3)	1 (0.4)	0.09
Other infections^a	8 (3.6)	4 (1.8)	0.22

^aWound infection, nephritis and erysipelas.

Figure 1 shows the distribution of HAI frequency according to month of occurrence. This suggests that the observed effect was similar over time and not an ongoing reduction.

Figure 1.

Frequency of hospital acquired infections by month in multi-bed and single-bed accommodation.

Figure 2 presents hazard ratios (HR) for the days to first HAI occurrence starting 48 hours after hospital admission. By time measured from admission to first HAI, the hazard ratio (HR) for single-bed accommodation was 0.62, 95% Confidence Interval (CI) 0.43–0.91, p = 0.01). Adjusting for age, sex, infection at admission, risk of sepsis, use of catheter, treatment with prednisone or methotrexate, and comorbidity index, did not change the result, at a HR of 0.65 (95% CI 0.45–0.95, p = 0.03) (Figure 2). None of the above-mentioned variables were found to be effect modifiers for HAIs. Infection at admission was associated with a reduced incidence of HAIs (HR = 0.33; 95% CI: 0.22–0.50). Of the catheter users 15% had a hospital-acquired urinary tract infection compared to 9% in the non-users (p = 0.11).

Figure 2.

Cumulative incidence of hospital acquired infection (HAI) in geriatric patients accommodated in multi-bed room versus single-bed rooms (y-axis presents the number of patients affected per 100).

Discussion

This study of a university hospital in Denmark found the incidence of HAIs to be significantly lower in geriatric patients accommodated in single-bed rooms compared to those in multi-bed rooms. This difference was largely driven by a lower rate of urinary tract infections in the single-bed room cohort.

Our findings are consistent with a Canadian study, which used a similar design, that is, looking at differences in infection rates among patients across a move from multi-bed rooms to single-bed rooms.¹⁹ It found a reduction in the incidence of nosocomial vancomycin-resistant Enterococcus and methicillin-resistant Staphylococcus aureus colonization and vancomycin-resistant Enterococcus infection although there was no change in nosocomial Clostridium difficile or methicillin-resistant Staphylococcus aureus infections. Similarly, Park et al. found in an epidemiological study that an increasing proportion of single-bed patient rooms in a hospital was significantly associated with fewer hospital-acquired methicillin-resistant Staphylococcus aureus infections.²⁰

There are many possible explanations for our findings. Aside from the potential protective effect of single-room accommodation leading to lower transmission of pathogens between patients, other factors need to be considered, too. These include improved ventilation of rooms in a new hospital building, additional space for performing procedures on patients, and a move away from shared bathroom and toilet facilities. O’Neill et al. found that the risk of central line-associated bloodstream infections was largely reduced by the use of single-bed room accommodation, which was explained by a single-bed room providing more space for performing procedures, reducing the risk of adverse outcomes.¹⁸

Our study showed that patients admitted with a recognized infection appeared to have a lower incidence of hospital-acquired infections compared to patients not admitted with an infection. This could be caused by a larger proportion of patients being treated with antibiotics when an infection is already present and recognized, thus providing a prophylaxis for various pathogens. As the most frequent infections at admission were urinary tract infection and pneumonia, this could reduce the probability of acquiring a HAI. Treatment with antibiotics is a well-known risk-factor for acquiring Clostridioides difficile infection,²⁴ although this infection was rare among our patient population.

Viral infections were not included in this study although single-bed accommodation might be expected to have the considerable impact on infection spread via the airborne route.

Strengths and limitations

The strength of the present study primarily rests in its design, with little indication of selection bias. The catchment area of the old and the new hospital were similar as were the admission criteria and staffing and clinical procedures. The frequency of HAIs varies with yearly seasons, peaking in winter,²⁵ and our study considered the same study period in terms of time of the year (autumn), with no recording of outbreaks of flu or other viral infections. Hand hygiene is widely regarded as the most important infection control measure. Poor access to sinks is associated with decreased hand washing,²⁶ but in our study, access to sink and hand sanitizing facilities did not differ between the two study cohorts. The difference in SIRS scores on admission between the two cohorts is notable, suggesting that patients in single-bed accommodation were generally sicker and, thus, at greater risk of infections. While our study suggests an association between single-bed accommodation and lower risk of acquiring HAI, we cannot comment on the causality.

Unfortunately, it was not possible to blind the data collectors for the two time periods that allowed us to distinguish multi-bed and single-bed accommodation, as the dates were visible in the electronic medical record. However, the key outcome data was presence or absence, thus lowering the risk of misclassification. The analyses were not adjusted for non-antibiotic prophylaxis of urinary tract infection. Further, there was a risk of underreporting infection incidence within two days after discharge since some patients declined follow-up by the geriatric team in their home. In an earlier study, we found that only 4% of patients did not receive a follow-up visit within three days after discharge.²¹ However, we can reasonably assume that rates of declining follow-up visits and resultant missing data were similar between the two patient cohorts. Finally, it is important to note that patients were transferred to the geriatric wards from different wards, most frequently the emergency department, the department of neurology, and the department of orthopaedics after 1 median day. However, time to transfer and incidence of infection within the first two days after the transfer were equal in the two cohorts.

Conclusions

In geriatric wards, single-bed accommodation might prevent hospital acquired infection, in particular urinary tract infections compared to multi-bed accommodation. The findings from this observational study call for further experimental research to determine the extent of prevention of HAI single-bed accommodation can offer. There is a more general need for further work on other potentially beneficial or harmful effects of single-bed accommodation to inform the future design of hospitals.

Footnotes

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Ethics approval

The authors declare that all the research meets the ethical guidelines.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Merete Gregersen

References

European Center for Disease Prevention and Control. Point prevalence survey of healthcare-associated infections and antimicrobial use in European acute care hospitals 2011–2012. Stockholm: ECDC, 2013.

World Health Organisation. Report on the burden of endemic health care-associated infection worldwide – a systematic review of the literature. Geneva: WHO, 2011.

Umscheid

Mitchell

Doshi

, et al. Esstimating the proportion of healthcare-associated infections that are reasonably preventable and the related mortality and costs. Infect Control Hosp Epidemiol 2011; 32: 101–114.

Jenkins

DR.

Nosocomial infections and infection control. Medicine 2017; 45: 629–633.

Department of Health. Health building note 04-01 supplement 1 – isolation facilities for infectious patients in acute settings. Report. UK: Department of Health, 2013.

Stiller

Salm

Bischoff

, et al. Relationship between hospital ward design and healthcare-associated infection rates: a systematic review and meta-analysis. Antimicrob Resist Infect Control 2016; 5: 51.

Voigt

Mosier

Darouiche

Private rooms in low acuity settings: a systematic review of the literature. Herd 2018; 11: 57–74.

Cepeda

Whitehouse

Cooper

, et al. Isolation of patients in single rooms or cohorts to reduce spread of MRSA in intensive-care units: prospective two-centre study. Lancet 2005; 365: 295–304.

Bracco

Dubois

Bouali

, et al. Single rooms may help to prevent nosocomial bloodstream infection and cross-transmission of methicillin-resistant Staphylococcus aureus in intensive care units. Intensive Care Med 2007; 33: 836–840.

10.

Bonizzoli

Bigazzi

Peduto

, et al. Microbiological survey following the conversion from a bay-room to single-room intensive care unit design. J Hosp Infect 2011; 77: 84–86.

11.

Teltsch

Hanley

Loo

, et al. Infection acquisition following intensive care unit room privatization. Arch Intern Med 2011; 171: 32–38.

12.

Lorenz

Dreher

HM.

Hospital room design and health outcomes of the aging adult. Herd 2011; 4: 23–35.

13.

Levin

Golovanevski

Moses

, et al. Improved ICU design reduces acquisition of antibiotic-resistant bacteria: a quasi-experimental observational study. Crit Care 2011; 15: R211.

14.

Ellison

Southern

Holton

, et al. Hospital ward design and prevention of hospital-acquired infections: a prospective clinical trial. Can J Infect Dis Med Microbiol 2014; 25: 265–270.

15.

Simon

Maben

Murrells

, et al. Is single room hospital accommodation associated with differences in healthcare-associated infection, falls, pressure ulcers or medication errors? A natural experiment with non-equivalent controls. J Health Serv Res Policy 2016; 21: 147–155.

16.

Stiller

Schroder

Gropmann

, et al. ICU ward design and nosocomial infection rates: a cross-sectional study in Germany. J Hosp Infect 2017; 95: 71–75.

17.

Darley

ESR

Vasant

Leeming

, et al. Impact of moving to a new hospital build, with a high proportion of single rooms, on healthcare-associated infections and outbreaks. J Hosp Infect 2018; 98: 191–193.

18.

O’Neill

Park

S-H

Rosinia

The role of the built environment and private rooms for reducing central line-associated bloodstream infections. PLoS One 2018; 13: e0201002.

19.

McDonald

Dendukuri

Frenette

, et al. Time-series analysis of health care-associated infections in a new hospital with all private rooms. JAMA Intern Med 2019; 179: 1501–1506.

20.

Park

S-H

Stockbridge

Miller

, et al. Private patient rooms and hospital-acquired methicillin-resistant Staphylococcus aureus: a hospital-level analysis of administrative data from the United States. PLoS One 2020; 15: e0235754.

21.

Kimberly

Dawn

Infection in an aging population. Curr Opin Microbiol 2016; 29: 63–67.

22.

Cairns

Reilly

Stewart

, et al. The prevalence of health care-associated infection in older people in acute care hospitals. Infect Control Hosp Epidemiol 2011; 32: 763–767.

23.

Pedersen

Gregersen

Barat

, et al. Early geriatric follow-up after discharge reduces readmission – a quasi-randomised controlled trial. Eur Geriatr Med 2016; 7: 443–448.

24.

McDonald

Sd Gould

Vital signs: preventing Clostridium difficile infections. Morbid Motal Week Rep 2012; 61: 157–162.

25.

Haller

Eckmanns

Benzler

, et al. Results from the first 12 months of the national surveillance of healthcare associated outbreaks in Germany, 2011/2012. PLoS One 2014; 9: e98100.

26.

Deyneko

Cordeiro

Berlin

, et al. Impact of sink location on hand hygiene compliance after care of patients with Clostridium difficile infection: a cross-sectional study. BMC Infect Dis 2016; 16: 203.