Hospital Staffing and Health Care–Associated Infections: A Systematic Review of the Literature

Hospital Staffing and Health Care–Associated Infections:
A Systematic Review of the Literature
Brett G. Mitchell, PhD, MAdvPrac; Anne Gardner, PhD; Patricia W. Stone, PhD, RN, FAAN; Lisa Hall, PhD;
Monika Pogorzelska-Maziarz, PhD
Background: Previous literature has linked the level and types of staffing of health facilities to the risk of acquiring a health
care–associated infection (HAI). Investigating this relationship is challenging because of the lack of rigorous study designs
and the use of varying definitions and measures of both staffing and HAIs.
Methods: The objective of this study was to understand and synthesize the most recent research on the relationship of
hospital staffing and HAI risk. A systematic review was undertaken. Electronic databases MEDLINE, PubMed, and the Cu-
mulative Index to Nursing and Allied Health Literature (CINAHL) were searched for studies published between January 1,
2000, and November 30, 2015.
Results: Fifty-four articles were included in the review. The majority of studies examined the relationship between nurse
staffing and HAIs (n = 50, 92.6%) and found nurse staffing variables to be associated with an increase in HAI rates (n = 40,
74.1%). Only 5 studies addressed non-nurse staffing, and those had mixed results. Physician staffing was associated with an
increased HAI risk in 1 of 3 studies. Studies varied in design and methodology, as well as in their use of operational defi-
nitions and measures of staffing and HAIs.
Conclusion: Despite the lack of consistency of the included studies, overall, the results of this systematic review demon-
strate that increased staffing is related to decreased risk of acquiring HAIs. More rigorous and consistent research designs,
definitions, and risk-adjusted HAI data are needed in future studies exploring this area.
Health care–associated infections (HAIs) are a serious
patient safety issue that result in increased morbidity
and mortality as well as excessive health resource utilization.1
Recent estimates from the United States show that on any
given day approximately 1 of every 25 inpatients in acute
care hospitals has at least one HAI. 2 In Europe HAIs also
represent a considerable burden, with more than 2.5 million
cases occurring each year, resulting in approximately 2.5
million disability-adjusted life years. 3 Given the significant
burden of HAIs with the potential for adverse outcomes in
patients, there is much interest in understanding their trans-
mission, prevention, and control. One particular issue is the
relationship between levels and types of staffing of health fa-
cilities and HAIs. A number of organizational factors that
influence the risk of HAIs have been identified, including
nurse-to-patient ratio, level of nurse education, and job type
(that is, temporary or permanent). 4–9 While nurses in par-
ticular are tasked with the responsibility of providing daily
bedside patient care, all health care workers are responsible
for applying infection prevention and control practices to
prevent HAIs. 4
Examining the association between hospital staffing
and HAIs is not without challenges as it requires the use of
standardized HAI case definitions, adequate data sources, and
complex risk adjustment methods. 9 Furthermore, the web
of causation linking staffing and HAI is difficult to under-
stand and may include factors such as the complexity of the
infection process, lack of time to comply with infection
control measures, and job-related burnout. 9 Methodologi-
cal issues in studies examining the association between hospital
staffing and adverse outcomes have also been identified. These
include lack of application of standardized definitions of nurse
staffing, different databases, and diverse risk adjustment
methods. 10 In addition, the temporal relationship between
staffing and HAI occurrence has recently been noted as a
methodological problem in studies examining hospital staff-
ing and HAI.11 HAIs are by definition infections that occur
48 hours after hospital admission. 11 Hence, staffing levels
should be examined about 48 hours prior to detectable in-
fection and not when the HAI is detected.
In 2008 a systematic review was undertaken to examine
the relationship between hospital staffing and HAIs. 12 The
researchers found that the majority of studies reported a sig-
nificant association between the nurse staffing variable(s)
studied and risk of HAIs. Since the completion of that review,
there has been a growing interest in infection prevention and
control research, particularly research into HAI prevention.13
Further, this previous review included articles published up
to 2007, and there have been numerous articles exploring
staffing and the risk of infection published since that date.
Understanding and synthesizing the most recent research on
1553-7250/$-see front matter
© 2018 The Authors. Published by Elsevier Inc. on behalf of The Joint
Commission. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
https://doi.org/10.1016/j.jcjq.2018.02.002
The Joint Commission Journal on Quality and Patient Safety 2018; 44:613–622

hospital staffing and HAI risk will inform health adminis-
trators, policy makers, and researchers on strategies for
preventing HAIs and thereby improving patient outcomes.
This systematic review therefore aims to examine the asso-
ciation between hospital staffing and patients’ risk of
developing HAIs in hospital settings.
METHODS
A systematic review of the literature was undertaken to iden-
tify publications that examine the relationships between
hospital staffing and patients’ risk of developing an HAI in
the hospital setting. The approach used is consistent with
a previous systematic review of this topic.12 Reporting of this
systematic review complied with the preferred reporting items
for systematic reviews and meta-analyses (PRISMA)
guidelines. 14
Protocol and Registration
The protocol for conducting this review was registered prior
to commencement of the review and can be accessed on the
international prospective register of systematic reviews
(PROSPERO) (registration number: CRD42015032398).
Search Strategy
A systematic search of the literature was conducted accord-
ing to the registered protocol. Electronic databases PubMed
and the Cumulative Index to Nursing and Allied Health Lit-
erature (CINAHL) were searched for studies published
between January 1, 2000, and November 30, 2015. The
search was performed on December 7, 2015. A combina-
tion of Medical Subject Headings (MeSH) and free-text terms
were used, including “infection control,” “staffing,” and
“healthcare associated infection.” For retrieved articles, a
manual search of the reference lists was also performed to
identify any additional studies. Searches were restricted to
studies published in the English language only.
Selection Criteria
The inclusion criteria were all observational studies (cohort,
case control, or cross-sectional) examining the relationship
between staffing and HAI in hospital settings. Random-
ized control trials were not available for inclusion. Given the
complex and multifaceted nature of HAI and the ethical con-
cerns about randomizing nurse staffing, conducting
randomized control trials on this topic is often not feasible.13
Studies describing health care staffing at the unit or orga-
nizational levels using variables such as level of staffing
(number of staff, staff-to-patient ratio, staff hours per patient-
day, years of experience, educational level), skill mix, use of
float or nonpermanent staff, absenteeism and/or overtime,
and workload were included. Exclusion criteria included grey
and non-peer reviewed literature, reviews, editorials, com-
mentaries or policy statements, articles reporting on
community-acquired infections, and articles written in lan-
guages other than English.
Definitions
For the purpose of this systematic review, the following defi-
nitions were used:
• Hospital staffing was defined as nurse staffing, medical
staffing, or infection prevention and control staffing.
• Nurse staffing levels were described using one or more
of the following variables: level of staffing (nurse-to-
patient ratio or nursing hours per patient-day or
admission), skill mix, use of float or nonpermanent staff,
absenteeism and/or overtime, workload.
• Health care–associated infections comprised blood-
stream infection, pneumonia, urinary tract infection,
wound or surgical site infection, organism-specific in-
fections (for example, Clostridium difficile infection) that
were defined as being health care–associated in the studies
included in the review. The definition of HAI in the
included studies was based on a recognized standard;
that is, a definition agreed on or published by a pro-
fessional association or government agency (for example,
the Centers for Disease Control and Prevention [CDC]),
a definition widely used in the published literature, or
an International Classification of Diseases, Ninth Re-
vision, Clinical Modification (ICD-9-CM) code that
constitutes an HAI (not just any infection).
Study Selection
The titles and abstracts of all articles identified were exam-
ined and assessed for relevance and appropriateness to the
systematic review aim, and those not relevant were ex-
cluded. The full texts of potentially relevant articles were
obtained to further assess eligibility based on the inclusion
and exclusion criteria. Articles with data relevant to the sys-
tematic review were included. The electronic database search
and study selection process were performed by trained re-
search assistants. At each stage of the study selection process,
10% of articles retrieved were selected at random and re-
viewed by the study lead author as a cross-check against study
eligibility. Any discrepancies in the application of the in-
clusion or exclusion criteria were resolved by the lead author.
In addition, any disagreements between study authors in
regard to the definitions were resolved by discussion with
all authors.
Data Extraction
A data extraction form in Microsoft Excel (Microsoft Corp.,
Redmond, Washington) was designed for the purpose of ex-
tracting data for the systematic review. For each eligible study,
the following data were extracted: author(s), year of publi-
cation, country of study, study design, study population, unit
of analysis (patient, unit, or hospital), sample, setting, staff-
ing category studied, staffing data source, staffing variables
and parameters, type of HAI, HAI definition, and HAI
614 Brett G. Mitchell, PhD, MAdvPrac, et al Hospital Staffing and Risk of Infection

incidence or prevalence data. Data extracted were crossed-
checked by a different research assistant.
Risk of Bias
An assessment of study quality and risk of bias in the ar-
ticles included in the review was conducted using the
Newcastle–Ottawa Scale.15,16 The content validity and inter-
rater reliability of this tool has been established.16 One reviewer
undertook this assessment independently, with a random 10%
of articles reviewed by a second reviewer. No discrepancies
were identified.
Data Analysis
Extracted data from included studies were synthesized and
summarized in evidence tables. Summary tables include
studies that examined nurse staffing and single site–specific
HAI, nurse staffing and multiple types of HAI, nurse staff-
ing and organism-specific HAI, nurse staffing and unspecified
HAI, and non-nurse staffing and HAI. Given the hetero-
geneity of the studies included in the systematic review,
pooling of data in a meta-analysis was not feasible.
RESULTS
Overview
The literature search yielded a total of 1,247 articles. After
excluding duplicates and reviewing the titles and abstracts,
there were a total of 90 articles remaining for full text review.
The 90 articles were screened against the study inclusion and
exclusion criteria. A total of 35 articles were excluded because
they were reviews, editorials, commentaries or policy state-
ments, or articles reporting on community-acquired infections.
Fifty-four studies met the inclusion criteria and were in-
cluded in the final systematic review (Figure 1).
Study Characteristics
Data on the characteristics of the 54 included studies are
presented in a supplementary table (Appendix 1, available
in online article). Of the 54 studies, 29 (53.7%) were un-
dertaken in the United States. Half the studies (n = 29;
53.7%) used a cohort (retrospective or prospective) or lon-
gitudinal study design. Analysis of data was performed for
most of the studies at the patient level (n = 28; 51.9%). The
most common HAIs studied were bloodstream infectionFlow Diagram of the Phases of the Systematic Review
Figure 1: This flow diagram provides the phases of article identification and selection, which resulted in the identification
of 54 articles that were deemed eligible for inclusion. Prepared in accordance with Moher D, et al. Preferred reporting
items for systematic reviews and meta-analyses: The PRISMA statement. J Clin Epidemiol. 2009;62:1006–1012.
Volume 44, No. 10, October 2018 615