Advanced use of electronic health record (EHR) tools that standardize care and increase patient access to their providers are purported to improve care quality and safety; however, clinician uptake of these advanced tools has been slow. As U.S. health care reform measures continue to encourage the use of teams in care delivery, team dynamics may play a role in innovation adoption decisions of clinicians. Little detail is known about whether clinicians' use of technological innovations is influenced by peer and team use patterns.

From the patient perspective, clinical trials are the primary mechanism by which new approaches to cancer treatment are evaluated; yet, only a small proportion of eligible cancer patients are offered the opportunity to participate in clinical trials, and fewer actually become enrolled. Trends toward more patient-centered care delivery have encouraged shared decision-making between patients and their care providers, which may influence patient awareness and consideration of medical innovations as treatment options, and ultimately decide to enroll in clinical trials.

This dissertation research assesses whether there are team effects associated with individual adoption of innovations, as well as whether there are normative pressures or internal motivations, as opposed to more passive interpersonal influences, that are related to clinician innovation adoption. It also assesses whether individual attitudes and awareness alone, or more active information sharing are more likely to result in patient innovation uptake.

Findings from self-reported clinician and patient surveys in an integrated delivery system suggest that interpersonal interactions contribute to individual clinician and patient awareness and eventual use of innovations in the clinical setting. Specifically, clinician adoption of advanced EHR tools was associated with their team members' use of the innovation. However, team cohesion was not directly associated with individual adoption and did not moderate the relationship between team use and eventual individual use of advanced tools. This suggests that more a passive transfer of knowledge may account for team influence on individual team member's adoption of advanced EHR tools. Conversely, medical oncologist behaviors and patient care experience, both play important roles in patients' ultimate decision to enroll in cancer clinical trials. This points to the need for clinicians to more actively engage with their patients, and greater patient understanding of trials to effectively encourage cancer clinical trial participation.

There are growing numbers of individuals with at least one chronic disease in the United States, creating a greater need for primary care services despite limited health care resources within safety net clinics. Patient self-management has been studied as a possible solution to improve the efficiency and effectiveness of primary care. In order to work, patients must feel confident in their ability to manage their own health and engage in the decision making process of their care. Providers also need adequate time to spend with patients to provide sufficient information as well as listen to patient concerns.

Three safety net clinics in Northern California underwent practice redesigns to enhance care for obese patients and improve their self-management skills. A mixed methods approach was used to assess the effects of the practice redesign on staff, clinicians, and patients. Patients were mailed questionnaires during early and late implementation of the practice redesign. Questionnaire measures included patient-provider communication (based on the CG-CAHPS), patient activation (PAM-13), general health (SF-12v2), physical activity, dietary habits, chronic conditions, and demographic information. Clinical outcome measures from clinic administrative data were merged with patient questionnaire data to evaluate patient outcomes relative to self-reported measures. Key informant interviews were conducted with randomly selected clinic staff and clinicians, also during the same baseline and six-month follow-up periods. Topics included practice change implementation experiences, team activities, interactions with overweight or obese patients, team development activities, and practice characteristics.

Higher patient activation was associated with increased odds of regular fruit and vegetable consumption. For physical activity, the presence of comorbidities affected the relationship between patient activation and physical activity. Although the patient-provider relationship had a strong bivariate relationship with patient activation, it did not impact the relationship between patient activation and health behaviors. As seen in previous studies, patient activation was related to systolic blood pressure, diastolic blood pressure and weight. However, the clinical practice redesign only had an effect on diastolic blood pressure. The implementation of a teamlet model within safety net clinics to improve patient health may be feasible, but the composition of the teamlet may need to be modified for financial sustainability. Future change efforts should have support from different levels of management and staff, preferably championed by multiple people to aid the sustainability of the change in the face of staffing changes.

Despite significant advances that have been made in breast cancer diagnosis and treatment, a disproportionate burden of the disease continues to fall on women from minority groups. Studies suggest that differentials in the quality of breast cancer care in the treatment stage influence breast cancer disparities. Because treatment is defined by medical intervention, the health system plays a critical role in understanding breast cancer disparities and needs to be further examined. Moreover, studies examining such differentials have focused on aggregated racial data, which may be masking smaller ethnic groups that may be at higher risk. This dissertation attempts to clarify the health system contribution to breast cancer quality of care disparities and highlights the need to examine breast cancer quality of care disparities using disaggregated racial/ethnic data in order to help inform the design of more targeted quality improvement efforts.

Paper #1 examines the relationship between designated specialty cancer centers and breast cancer quality of care. In cancer care delivery, there has been an emerging trend for cancer centers to obtain special designations from one or more of the following institutions: the National Cancer Institute (NCI), the National Comprehensive Cancer Network (NCCN), and the American College of Surgeon’ Commission on Cancer. These designations are markers for high quality cancer care, yet little is known about the actual effectiveness of the care they provide or the communities that they tend to serve.

Paper #2 examines breast cancer quality of care differences between Asian women (aggregated and disaggregated) and non-Hispanic White women. Cancer disparities research often focus on examining communities in aggregate. Recently, more scholars are utilizing disaggregated ethnic data when assessing cancer disparities in order to acknowledge the diversity of these large communities. Little is known about breast cancer quality of care disparities among Asian American ethnic groups.

Paper #3 examines the relative contribution of within hospital and between hospital effects on racial/ethnic disparities on the quality of breast cancer treatment. Existing literature posits that there are two mechanisms at the health systems level that contribute to disparities in the quality of care received by minority groups. Some studies suggest that disparities are due to minorities receiving differential treatment within the same institution as White patients, while others propose that minorities tend to receive care from lower quality health care settings. Studies to date have not examined the ways in which the health system may be contributing to breast cancer quality of care disparities.

Data from the California Cancer Registry (CCR) was used for the analyses. The study population consisted of women ages 20 years or older and who have had a first invasive primary breast cancer diagnosis between 2011-2014.

This dissertation contributes to the existing literature on breast cancer disparities by clarifying the mechanism of the health system contribution to breast cancer disparities and uncovering disparities in disaggregated ethnic groups. Findings from this dissertation will provide important information for the development of quality improvement efforts in cancer care delivery.

Over half of all births in Mexico are attended in the public health care system. Prenatal care for these women is overwhelmingly at the primary healthcare clinics staffed by general practice physicians, medical residents, and in a few instances, professional midwives. Access to specialty care for those women who need it is dependent on the effective functioning of the referral system.

This dissertation seeks to understand the functioning of referral systems, the quality of care provided to women with obstetric-related referrals and aspects of care coordination to facilitate care of referred patients using a mixed methods study of a network of public health primary care clinics and secondary-level hospitals in Oaxaca and Guerrero, two of the poorest Mexican states.

Chapter 1 found that having insurance coverage, in this case, the public health insurance coverage Seguro Popular, led to greater odds of referral and therefore presents a policy solution to help break down barriers to accessing specialty care. Chapter 2 assessed the quality of care delivered in the prenatal period, measured as the proportion of screenings and preventive care actions carried out. No significant difference in the quality of care delivered was observed between referred and non-referred women. Chapter 3 assessed the change in visit-level quality of care pre- and post-referral for referred versus non-referred women. We found significant declines in quality of care over time in both referred and non-referred women, with referred women experiencing a significantly greater decline than their non-referred counterparts. Qualitative data highlight areas of process losses and potential gains with referred women and shared prenatal care between primary care providers and specialists.

While the referral system elements of physical infrastructure and human resources tend to be a focus of initiatives to strengthen systems in resource-constrained settings, failing to strengthen care coordination hamper the development of strong referrals systems that can successfully care for women who require access to more specialized care.

Health information technology (IT) has become an increasingly common part of the U.S. healthcare system. This three-paper dissertation examines barriers, facilitators, and correlates of health IT adoption and impacts of health IT use in several important settings.

The first paper examines the impact of electronic warnings in diverting non-indicated prescriptions in ambulatory settings. Specifically, we measured the incidence of antibiotic prescriptions for acute bronchitis and upper respiratory infections using a nationally-representative dataset. We found evidence that, overall, electronic warnings reduce the likelihood of antibiotic prescription receipt by about 20%. However, despite the recent increase in use of electronic warnings, antibiotic prescribing is not on the decline, suggesting that electronic warnings alone may not be sufficient to eliminate non-indicated prescriptions.

The second paper provides the first cross-year comparison of the use of electronic health records (EHRs) at local public health departments (LHDs) in the U.S. Using 2005 and 2010 data from the National Association of City and County Health Officials, we found that EHR usage has remained relatively steady at local public health departments. There is substantial churn, however, with approximately one-quarter of the sample adopting EHRs and another fifth of the sample discontinuing use of EHRs. Our study suggests that EHRs are not diffusing throughout LHDs as they are in other healthcare settings. Our results highlight departmental characteristics under which EHRs are commonly used and suggest potential places, such as poor or rural areas, where use of EHRs may be lower than expected.

The third paper examines participation in health information exchange (HIE) in two settings where HIEs have proven slow to diffuse--smaller-sized physician practices and federally qualified health centers (FQHCs). We conducted key informant interviews with stakeholders at practices and clinics. Our results suggest barriers to HIE adoption that exist at three levels--regional (e.g., existence of other area-level exchanges; number, type, and size of partner organizations), inter-organizational (e.g., strong relationships with exchange partners; achieving a critical mass of users), and intra-organizational (e.g., type of electronic medical record used; integration into organization's workflow). While some of these factors may be modifiable by health care organizations, limited solutions to overcome these barriers currently present a major challenge to the broad and effective use of HIE.

Health care delivery organizations are increasingly being encouraged to implement strategies to engage patients in their health and health care yet have varying capabilities to support patient-centered care. Improving the quality of health care delivery requires a better understanding of the multi-level organizational capabilities that enable patient engagement. This dissertation develops a conceptual framework to delineate the health care system, physician practice, and team factors associated with patient engagement in chronic care.

The first study links clinical data to surveys of patients with diabetes in fourteen community health center sites to examine if the availability of team member expertise is associated with patient experiences of chronic care, and whether the relationship is consistent for small versus large sites. The second study uses a nationally representative survey of physician practices and health systems to clarify how health systems may influence the adoption of patient engagement and chronic care management capabilities in constituent physician practices. The third study utilizes Mokken scale analysis and a survey of adult primary care practices to reveal potential pacing and foundation setting in the adoption of patient engagement strategies.

Objective

Data sources

Study design

Principal findings

Conclusions

Health care delivery and payment reform have intended to redesign care in a manner that is higher quality, better coordinated, and more efficient. As part of this shift, delivery systems are increasingly organizing in new ways, often moving toward greater organizational integration. One model that has emerged is the Clinically Integrated Network, or CIN, a set of independent provider organizations such as hospitals and physician groups that collaborate to provide high-value care. This dissertation examines antecedents and outcomes associated with hospital participation in a CIN.

The first study uses national survey data to identify the organizational characteristics and other foundational collaboration arrangements associated with hospital participation in a CIN. The second study evaluates CIN processes, using a survey of hospital activities to ask whether CIN-affiliated hospitals offer more integrated care delivery than hospitals not affiliated with CINs. The third study uses inpatient hospitalization data to evaluate care coordination outcomes associated with hospital participation in a CIN by testing for an association between CINs and potentially preventable inpatient utilization. Taken together, these studies provide the first national empirical evidence describing CIN activities, and the findings can help inform decisions of CIN leadership and regulatory authorities.