BACKGROUND/RATIONALE:
In its critique of the US medical system, the Institute of Medicine (IOM) endorsed patient-centered care (PCC) as one of the Six Aims of ideal health care. For over 12 years, the VA has assessed patients' perceptions of their health care using the patient-centered care surveys developed by the Picker Institute. Concurrently, another major force has been the movement toward evidence-based practice (EBP), treatment guidelines, and the collection and release of standardized outcome and performance measures. Here too the VA has been in the forefront, implementing a sophisticated performance measurement system using standardized metrics that make comparison with the private sector possible. There can be tension between these two movements. Is it possible to provide high quality medical care that is both consistently evidence-based and patient-centered? When evidence-based patient-centered care is delivered, what are the organizational factors that support such a potentially powerful combination?

OBJECTIVE(S):
The overall objectives of the proposed study are:
1. To identify the key hospital organizational characteristics that support the delivery of high quality, patient-centered inpatient care
2. To then understand in detail the behavioral, policy and process mechanisms whereby those organizational characteristics exert their influence.

METHODS:
The study involved a mixed methods design including the analysis of various secondary quantitative data and qualitative primary data collection.

For the quantitative study, we assembled an individual level working database for the cohort of patients hospitalized during FY04 who were (a) selected to contribute to a VA clinical performance measure through the external peer review (EPRP) chart review program, and who also (b) completed a patient satisfaction (SHEP) survey regarding their experiences during the index EPRP hospitalization. We focused on three patient groups: acute myocardial infarction (AMI), congestive heart failure (CHF), and community-acquired pneumonia (CAP). To the individual records for these patients we appended aggregate organizational culture and employee satisfaction scores based on AES respondents from inpatient medical units at each facility. Self-reported demographic characteristics were also available from the SHEP. Descriptive statistics, bivariate analyses, and hierarchical linear modeling were used to examine the relationships between culture, clinical processes, and patient-centered care. The hierarchical (mixed) modeling allowed us to account for the use of data at both the individual and facility levels.

For the qualitative phase, 142 interviews with the leadership and staff at 12 VA facilities were analyzed using QSR qualitative software. Inter-rater reliability was established and the research team coded interviews from multiple sites, using a code book that was developed using the tenets of grounded theory. Codes were elaborated and expanded in an iterative fashion to capture new concepts as coding proceeded.

FINDINGS/RESULTS:
We organized the quantitative data analysis around two groups of variables: patient characteristics and organizational characteristics. Regarding the former, age was negatively associated with adherence to EBP but positively associated with adherence to PCC, for both AMI and CHF. Age was not a significant predictor in the combined model predicting the average of both outcomes (EBP, PCC), probably reflecting an interaction. For CAP patients, age was positively associated with adherence to EBP, PCC, and the combined score.

Health status was unrelated to EBP for both cardiology conditions, but was positively associated with CAP EBP. However, health status was associated positively with PCC and the combined score for all three diagnostic categories.

Minority status was not associated with adherence to EBP for any diagnostic category. However, among CHF and CAP patients, there was a significantly lower adherence rate for PCC and for the combined outcome measure for non-whites.

Turning to the organizational factors, none were significant predictors of EBP for AMI patients. For CHF patients, individual feedback, working at a fast pace, and bureaucratic culture were associated positively with high performance. For CAP patients, staff satisfaction with the intrinsically-rewarding aspects of their work was positively associated with greater adherence to EBP.

For PCC, staff intrinsic satisfaction was a significant positive predictor for all three diagnostic groups, and fast pace of work was positive and significant for CHF and CAP patients. Group culture was negatively associated among CHF and CAP patients. Finally, bureaucratic culture was negatively associated for the AMI patients.

For the combined model, a positive association with intrinsic satisfaction and a negative association with group culture were significant for CAP patients only.

Systematic analysis of the qualitative data identified several differences between facilities that scored high on both EBP and PCC measures and facilities that scored low on both EBP and PCC measures. High and low scoring facilities frequently cited similar barriers (e.g. lack of resources, competing priorities) and facilitators (e.g. education, communication) to practicing care that is both evidence-based and patient-centered. However, in examining these barriers and facilitators several differences between high-performing and low-performing facilities emerged.

High-performing facilities had an active, innovative improvement culture in which individual accountability and staff engagement in problem solving were cited. Low-performing facilities had a passive, punitive culture, in which lack of individual accountability, a culture of blame, resistance to change and institutional burnout due to organizational factors and structures that are perceived as unchangeable were common.

Providers in high-performing facilities had autonomy and institutional support to provide care that is patient-centered, while not being fully constrained by VA performance measures and evidence-based guidelines that may not be appropriate for particular Veterans. Providers in low-performing facilities saw bureaucratic systems and organizational structures within the VA as factors that negatively impacted their ability to deliver care that is both EBP and PCC.

High-performing facilities took multidisciplinary approaches to care in which all members of the team were of equal importance and communication was open and multidirectional between all levels of the organization, including communication with the Veterans. In contrast, low-performing facilities had more structured approaches to multidisciplinary care and communication was more formalized and sometimes strained between various levels of the organization, including communication with the Veterans.

IMPACT:
The project generated knowledge about organizational characteristics that can be assessed and strengthened to facilitate patient-centered care models.

---

External Links for this Project

---

PUBLICATIONS:

Journal Articles

1. Cleary PD, Meterko M, Wright SM, Zaslavsky AM. Are comparisons of patient experiences across hospitals fair? A study in Veterans Health Administration hospitals. Medical care. 2014 Jul 1; 52(7):619-25. [view]
2. Austin EL, Pollio DE, Holmes S, Schumacher J, White B, Lukas CV, Kertesz S. VA's expansion of supportive housing: successes and challenges on the path toward Housing First. Psychiatric services (Washington, D.C.). 2014 May 1; 65(5):641-7. [view]

Conference Presentations

1. Engle RL, Holmes SK, Mohr DC, Seibert MN, Leyson J, Afable M, Meterko MM. Evidence-Based Practice and Patient-Centered Care: Doing Both Well. Paper presented at: AcademyHealth Annual Research Meeting; 2015 Jun 15; Minneapolis, MN. [view]
2. Engle RL, Holmes SK, Nealon Seibert M, Leyson JS, Afable MK, Meterko MM. Evidenced Based Practice and Patient-Centered Care: Doing Both Well. Paper presented at: VA HSR&D / QUERI National Meeting; 2015 Jul 10; Philadelphia, PA. [view]

DRA: Health Systems
DRE: none
Keywords: Outcomes, Research method, Satisfaction (patient)
MeSH Terms: none