Dissertations and Theses

Date of Degree


Document Type


Degree Name

Doctor of Philosophy (Ph.D.)


Health Policy and Management


Elizabeth Eastwood

Committee Members

Elizabeth Eastwood

William Gallo

Marianne Fahs

Subject Categories

Bacterial Infections and Mycoses | Geriatrics | Health Services Research | Nursing | Patient Safety | Public Health


Hospital nursing practice, infection prevention, patient safety, geriatrics, health services research


Background: Health care–associated infections, resulting from treatment received for medical or surgical conditions in a health care setting, represent a critical public health and patient safety issue, exacting substantial medical, social, and economic costs. The costliest among the leading causes of preventable health care-associated infections is central-line associated bloodstream infections (CLABSI), to which older adults (age 65 years and older) are particularly susceptible, especially during intensive care unit (ICU) stays. A rich body of research has empirically linked the quality of the nursing practice environment (NPE) in hospitals to both positive and negative patient outcomes; yet, surprisingly few studies have sought to examine relations between the hospital NPE and older adult CLABSI outcomes. This study aimed to fill this gap through analysis of de-identified data from the 2011 national Prevention of Nosocomial Infections and Cost-Effectiveness Refined (PNICER) study, provided by the Columbia University School of Nursing. PNICER was a three-year, mixed-methods study aimed at assessing infection prevention efforts at eligible National Healthcare Safety Network hospitals. In the present study, analyses included the following data from 739 PNICER participating hospitals: self-reported data on the organizational work climate—a measure of the NPE—from 1,665 hospital infection preventionists, most of whom were nurses, and data on ICU CLABSI occurrence among 19,383 Medicare patients. Specific research aims included:

AIM 1: Investigate the construct validity and test latent constructs of two healthcare organizational work climate instruments—the Leading a Culture of Quality Instrument for Infection Prevention (LCQ-IP) and the Relational Coordination Survey (RCS)—across PNICER-participating hospitals using Exploratory Factor Analysis (EFA).

AIM 2: Investigate whether LCQ-IP and RCS instrument items are predictors of hospitalized older adult ICU CLABSI outcomes across PNICER-participating hospitals using multivariate logistic regression.

AIM 3: Employ mediation analyses to examine whether LCQ-IP and RCS constructs mediate relations between hospital characteristics and older adult ICU CLABSI outcomes.

Methods: Aim 1 analyses involved Exploratory Factor Analysis (EFA) to investigate the construct validity of two PNICER healthcare organizational work climate instruments, the LCQ-IP and RCS. Two LCQ-IP EFA models were run, one including 20 unit-level variables, and one including seven individual-level variables, and an EFA was run on all 28 RCS items. Aim 2 analyses involved multivariate logistic regression to explore relations between eight organizational work climate domains—based on LCQ-IP and RCS factors identified during Aim 1—hospital characteristics, and older adult ICU CLABSI occurrence. A stepwise series of models were run that included various work climate and covariate permutations to assess the relative impact on CLABSI occurrence. Finally, building on findings from Aim 2, Aim 3 analyses employed the Joint Significance Test of mediation to investigate whether three organizational work climate domains identified during Aim 1—Quality Prioritization, Personal Satisfaction, and RN Relational Coordination—mediated the relations between select structural hospital characteristics and older adult CLABSI occurrence in order to obtain greater insight into how associations between these variables of interest operate. All analyses were conducted in SAS® 9.4.

Results: Exploratory Factor Analyses confirmed the construct validity and reliability of both the LCQ-IP and RCS instruments, capturing four-factor solutions for each. The four LCQ-IP factors included Psychological Safety, Quality Prioritization, Leadership and Change Orientation, and Personal Satisfaction, with acceptable internal consistency (Cronbach’s α = 0.909). The four RCS factors included Hospital Administration Relational Coordination, Environmental Services Relational Coordination, Physician (MD) Relational Coordination, and Bedside nurse (RN) Relational Coordination, also with acceptable internal consistency (Cronbach’s α = 0.768). Select organizational work climate domains were found to be statistically significant with CLABSI during multivariate logistic regression analyses—namely, Quality Prioritization, Personal Satisfaction, and Bedside Nurse Relational Coordination—although the magnitude and direction of those associations varied. Numerous structural hospital and infection prevention program department and policy covariates were found to impact the outcome, as well. However, goodness of fit statistics indicated that overall model fit was poor, with the fully adjusted and pruned regression models only explaining 5% of the variance of the CLABSI outcome variable, suggesting that indicators not assessed during the present study, such as patient and nurse-related characteristics, play an influential role in the work climate-CLABSI causal pathway. Results from Joint Significance Tests of mediation did not confirm the presence of mediation.

Conclusion: This research provided critical insight into the associations between elements of the hospital NPE and CLABSI among hospitalized older adults, thus contributing to the public health, nursing, and gerontological literature. The LCQ-IP and RCS EFAs validated the psychometric properties of these instruments, indicating their utility for researchers and providers seeking to assess the quality of the hospital organizational climate related to infection prevention. To the researcher’s knowledge, this was the first investigation of the PNICER LCQ-IP that demonstrated the construct validity of the full 27-item scale, which may enhance its variability and sensitivity, as well as improve its theoretical structure since identified constructs are more fully represented. RCS EFA findings further validated that this tool has utility in addressing core elements of interdisciplinary practice in hospitals surrounding infection prevention efforts, especially in regard to collaboration and communication.

The inconsistent findings related to organizational work environment-CLABSI associations that were observed during multivariate logistic were unexpected given that such a robust body of literature has documented the influential role that hospital work environment factors have on patient health care-associated infections, as well as that prior EFAs confirmed the psychometric properties of the LCQ-IP and RCS employed during PNICER. However, since goodness of fit tests revealed poor model fit, findings should be interpreted with caution. Additional research is needed to probe the nuanced associations between structural, organizational, and individual-level determinants of older patient safety and quality outcomes, such as CLABSI.

Results from the Joint Significance Test of mediation did not provide enough evidence to reject the null hypothesis and confirm a mediating effect among any of these factors. Despite these findings, the analyses offer important theoretical and scientific contributions to the nursing and patient safety literature. As structural hospital characteristics are often incorporated into analyses for descriptive purposes or as covariates in regression analyses but rarely as explanatory variables, this study’s inclusion of these variables as key predictors provides a foundation for future scientific inquiry into how such factors impact patient safety outcomes. Additionally, to the researcher’s knowledge, this was the first study to examine the hypothesized mediating role of the NPE on the hospital characteristics-older adult CLABSI outcome causal pathway. Despite its infrequent use in nursing research, mediation analyses, such as these, enable researchers to ask and answer more nuanced and arguably more meaningful research questions that extend beyond how one variable influences a particular outcome, thereby advancing the scope of scientific inquiry.

Findings from this study raise important questions surrounding what dimensions of the NPE are most consequential for older adult healthcare-associated infection occurrence in the ICU and how these NPE dimensions manifest within pre-existing hospital structures to shape patient outcomes. Additionally, analyses point towards the important yet overlooked role of nurse and patient characteristics in both shaping NPEs and driving outcomes. The complexity of these associations reinforces the need for further scientific investigation in this area, which should include advanced quantitative statistical methods aimed at teasing apart the nuance of these structural, organizational, and individual-level factors, in addition to qualitative methods, to provide important contextualization to the dynamics at play between these constructs. Such insights will help to highlight key areas for clinical and practice intervention to redress negative older adult patient outcomes and promote enhanced cultures of safety in ICUs and across hospital wards.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.