Dissertations and Theses

Date of Degree

9-30-2016

Document Type

Dissertation

Degree Name

Doctor of Public Health (DPH)

Department

Health Policy and Management

Advisor(s)

Alexis Pozen, PhD

Committee Members

Diana Mason, PhD, RN, FAAN

Diana Romero, PhD

Shoshanna Sofaer, Dr.P.H.

Subject Categories

Public Health

Keywords

Patient Engagement, Patient and Family Engagement, Patient Safety, Patient-centered Care, Patient Experience of Care, Patient Advocacy

Abstract

Introduction

Patient engagement involves the behaviors of patients, family members, and health professionals (i.e., doctors, nurses, and other healthcare staff) in a collaborative partnership to improve health and healthcare. It also constitutes organizational structures policies and procedures designed to foster and promote the active inclusion of patients and family members in health services delivery. Patient engagement is associated with enriched patient experience, patient safety and clinical effectiveness. An evidence base is essential for its translation from a conceptual framework to tangible programs that can be pragmatically implemented in healthcare delivery systems. However, little is known about how individual level factors, healthcare staff behaviors within hospital work systems and hospital characteristics support engagement, influence patients’ engagement comfort, and the impact on patients’ experience of care.

The intent of this doctoral dissertation is to perform a secondary analysis of the Consumers Union Patient and Family Engagement Survey to examine three key research questions:

  1. Which aggregated patient and hospital characteristics are associated with hospital staff behaviors that support patient engagement?
  2. Which aggregated patient and hospital characteristics are associated with patients’ level of comfort with engagement around safety issues during their hospital stay?
  3. What are the effects of hospital professional patient engagement behaviors and patients’ level of comfort with engagement around safety issues during their hospital stay on the patients’ experience of care?

Literature Review

The Multidimensional Framework for Patient and Family Engagement in Health and Healthcare, developed by Carman et al., delineates the levels at which engagement can and should take place. This includes direct care, organizational design and governance across an engagement continuum that includes consultation, involvement as well as partnership and shared leadership. Patient engagement behaviors are rooted in self-efficacy (i.e., activation) but also in perceived norms and consequences. They are also influenced by organizational, socio-demographic, socio-cultural, and task-related factors. Patient experience of care is shown to be positively associated with varied degrees of technical quality, patient safety, clinical effectiveness, patient-provider communication and treatment adherence. This doctoral dissertation focuses on the direct care level of engagement and studies the experiences, attitudes and patient engagement comfort of recently hospitalized patients.

Methods

I completed a secondary analysis of the probability-based, nationally representative Consumers Union Patient and Family Engagement Survey final data set. The study sample included non-institutionalized individuals’, age 25 to 75 residing in the United States who had at least one hospitalization in the last six months. I used the weighted sample of individual respondents in the secondary analysis to represent the population of interest in the United States. I used exploratory factor analysis as a data reduction technique to examine outcomes of interest. I explored the underlying factor structure of the 22 observed variables (i.e., 11 hospital staff behaviors and patients’ engagement comfort with those same 11 behaviors), and the correlation between the observed variables and the factors (i.e., factor loadings). I analyzed factor loadings, the proportion of variance explained by each factor, and any common factors. Principal component analysis was used to assess the amount of variance in the observed hospital staff behaviors and patient engagement comfort variables.

Outcome variable selection included hospital professional pain assessment and patient identification behaviors, patient engagement comfort with pain assessment and patient identification, and patients’ likelihood to recommend the hospital to a family member or friend. Logistic and ordinal logistic regression analyses were used to assess outcome variable relationships while controlling for each independent variable of interest. Complex survey data techniques and unvariate, bivariate and regression analyses were completed using SAS software.

Results

The sample was fairly evenly split between male (47%) and females (53%). The mean age was 53 and median was 50 years. Greater than 60% of respondents were White, non-Hispanic, 53% were married. Most were hospitalized in metropolitan, non-teaching hospitals with greater than 100 beds. Greater than 50% were hospitalized for a surgical admission and 56% had more than one previous hospitalization. The majority of respondents reported hospital professionals performed pain assessment and patient identification rather routinely. Higher proportions of these behaviors were reported by older patients, age 45 years plus, white, non-Hispanic, married, college educated patients who had advocate support during hospitalization. Logistic regression analysis showed that having a high school education was associated with a 0.31 odds of reporting pain assessment hospital engagement compared to those with a Bachelor’s degree or higher (p=0.02). Being admitted for a medical admission was associated with a 0.37 odds of reporting pain assessment hospital engagement compared to those admitted for surgery (p=0.02). These results were statistically significant in the logistic regression model (p<0.05).

Most respondents indicated they were comfortable with patient engagement. Greater than 80% of patients reported they were either very comfortable or comfortable with pain assessment and patient identification engagement. Most were older age, 45 plus years, married, and college educated. Ordinal logistic regression analysis showed having three previous hospital admissions was associated with a 0.55 odds of reporting feeling comfortable with pain assessment engagement compared to those with four or more previous hospital admissions (p=0.04). Results were statistically significant (p<0.05). The following results were also statistically significant in the patient identification ordinal logistic regression model. Being female was associated with a 0.55 odds of reporting feeling comfortable with patient identification engagement compared to males (p=<0.01). Having improved health status was associated with a 1.89 odds of reporting feeling comfortable with patient identification engagement compared to those whose health status worsened in the past year (p=0.05).

Greater than 80% of patients indicated they would definitely recommend the hospital to a family member or friend. The majority of these patients were older, 55 plus years, married, had advocate support during hospitalization, and had not experienced a medical error during their most recent hospitalization. Ordinal logistic regression analysis showed not having experienced one or more medical errors during the index (i.e., most recent hospitalization) was associated with a 2.00 odds of reporting they would recommend the hospital to a family member or friend compared to patients who reported they had experienced one or more medical errors during their most recent hospitalization (p=<0.01). Results were statistically significant in the ordinal logistic regression model (p<0.05).

Conclusion

This doctoral study attempted to fill an evidence gap around factors that influence comfort with patient engagement, measurement of hospital professional behaviors that support engagement and their impact on the patient experience of care. Study findings provide supporting evidence that patents perceive hospital professionals often proactively engage patients in pain assessment and patient identification. The analysis reinforced the role socio-demographic factors play in patient engagement. Logistic regression analysis showed that having a medical (i.e., non-surgical) hospital admission was positively associated with patient’s comfort with pain assessment engagement. Having only a high school education was negatively associated feeling comfortable with pain assessment engagement. With regard to gender, female and improved health status in the past 12 months, were positively associated with patient’s comfort with patient identification engagement. It also confirmed the statistically significant medical error experience on patients’ overall experience of care. Findings also support that patients’ admission type matters, particularly to pain assessment. Ultimately these findings can be used to identify opportunities to improve the nature of the patient-health professional relationship toward greater degrees of collaboration and engagement, particularly related to safety and quality of care. It addresses limited evidence as to what precisely patients’ feel comfortable doing when it comes to patient engagement. Future patient engagement research should focus on actual patient engagement behaviors, characteristics of engagement and continued evidence around measure development.

Available for download on Wednesday, January 31, 2018

Included in

Public Health Commons

Share

COinS