FALLING THROUGH THE CRACKS: Methodology
This survey research, designed by the Perinatal Survey Team, is descriptive research with primary data gathered from a purposive sample of 32 community agencies, organizations, and hospitals in Ramsey and Hennepin counties in the state of Minnesota during the years 1999-2001. Although data were collected at various times over a 28-month period, the study is basically cross-sectional rather than longitudinal, since it looks at agency policies and procedures as they existed during a specific window of time (1999). Subsequent data were collected solely to amplify and clarify the original data rather than to assess change over time. Qualitative methods were used to gather and analyze the data. Limited, descriptive, quantitative data were also gathered to support the qualitative analysis.
Qualitative methods were selected as the primary strategies for data gathering and analysis in this study, as this approach is the most suitable means of addressing the stated goals and objectives: describing the existing baseline of perinatal care coordination; determining interagency transactions in the delivery of such care; defining a model system of perinatal care; and explaining the current reimbursement system for perinatal care. Inductive methods, such as those employed in qualitative research, focus on obtaining a broad base of detailed knowledge about the subject under study. As such, they lend themselves well to open-ended and/or semi-structured questionnaires and successive waves of data gathering.
Unlike mathematical and statistical procedures which form the backbone of quantitative research and which tend to produce aggregate numeric data patterns, qualitative methods strive to capture the big picture in narrative terms. Also, qualitative methods can give the intricate details of phenomena that are difficult to convey with quantitative methods (Strauss & Corbin, 1990, p. 19). These details are then honed to converge on interpretation of the selected social problem(s).
Furthermore, one of the strengths of qualitative analysis is its focus on . . . naturally occurring, ordinary events in natural settings, so that we have a strong handle on what real life is like(Miles & Huberman, 1994, p. 10). Such is the intent of this research: to look carefully and objectively at the real world of existing perinatal care coordination. Qualitative analysis is a descriptive process, whichif carried out thoroughly and meticulouslyis the most effective route to dealing with the research question set forth in this study.
The survey instrument developed for use in this study was based on the previously described theoretical model developed by Keller et al (1998). Key components of the model and the survey include:
1). Mission, Philosophy, Scope of Practice
2). Community and Systems Level Activities
3). Client Level Activities
4). Financial and Capacity Issues
5). Comments and Recommendations
Within these categories were three types of questions: structured, semi-structured, and open-ended, with the latter two types comprising the bulk of the questionnaire (See Appendix 1). To develop the social, behavioral, and medical risk assessment components of the survey, the Minnesota Pregnancy Risk Assessment Form (MPAF) was used (See Appendix 2).
The basic survey was constructed and presented in two slightly different formats: the original community health center/public health nursing version and the adapted version used with health plans. Both survey instruments included the same content and questions; however, the language in the adapted version was modified slightly to render it applicable to health plans. Neighborhood Health Care Network (NHCN) staff reviewed the instrument, encouraged their Network members to respond, and made suggestions for revisions.
In addition, a shorter survey instrument was designed for gathering information from hospitals based on questions raised at Project LID Task Force meetings (See Appendix 3). Depending upon the structure of the organization being surveyed, either the health plan instrument or the hospital instrument, or both,were used. Hospital and health plan responders were instructed to use the survey as a guide, answering the questions where appropriate but providing additional information and narrative where the survey did not effectively ask questions about their perinatal care coordination activities.
All data were gathered by the Perinatal Survey Team, which was staffed by the perinatal research scientist, the perinatal and womens health consultant, and the infant mortality consultant from MDH, as well as the health systems program analyst and Project LID intern from the Minneapolis Department of Health and Family Support (MDHFS). Initially, the Team met with key staff from the other local public health agencies, and with representatives of the NHCN, community health centers, and public health nursing agencies to seek their advice on constructing appropriate objectives for the survey. These agencies were selected because they comprise safety net organizations for low-income and underserved populations, and consequently they serve a disproportionate number of high-risk women and infants.
The resulting survey was completed in writing and returned to the team by 15 of the 18 community health centers, as well as the three public health nursing agencies serving the two counties. These organizations were reimbursed for their participation in the survey process. In addition, each organization was offered an in-person, followup interview with a member of the team to review the survey forms and clarify any questions they might have.
Staff contacted the five health plans and twelve hospitals serving perinatal clients in the two counties and requested information on their perinatal case management and/or perinatal care coordination services. Copies of the surveys were faxed, and the contacts were told that they could either (1) fill-out the surveys and send or fax them back; (2) meet with one or more of the researchers to discuss the content of the surveys in addition to or in lieu of writing their answers; or (3) speak with one of the researchers over the phone to give answers and information.
All five health plans and nine of the twelve hospitals provided information for this study. Some health plans provided limited responses to the survey questions. Within a health plan or health system, there were many individuals and/or departments potentially involved in each element of the survey, and it was not always clear who these contact persons were. Also, the health plans and hospitals were not reimbursed for their participation, which may have influenced the time and effort they could devote to completing the survey. After reviewing an initial draft of this report, health plans provided additional information to complete the data they believed was missing from the original survey responses.
Analysis of the data was an iterative process, using an inductive method. Each successive wave of data gathering broadened the scope of the analysis and at the same time refined the focus of inquiry. Data gathering continued until no new themes emerged and all questions raised by earlier data were clarified and resolved.
Initially, each team member reviewed a group of surveys individually, using a guide (see Appendix 4) designed to extract and summarize the key themes. Subsequently, the team discussed and revised individual summaries as a group, and then prepared overall summaries of responses from each type of organization. Through discussion, additional questions were generated, and the need for further clarification of responses was evident. Many phone and face-to-face follow-up sessions were conducted by the survey team, which resulted in additional waves of data.
Survey responses from the community health centers and public health agencies indicated that the systems in which they worked had changed significantly in recent years, negatively impacting their ability to provide services. Responses also pointed to major demographic changes in the community that were adding complexity to their work. These issues led the team to an historical analysis of the way in which Minnesota had implemented Medicaid reform in the mid- to late-1980s during the federal Medicaid expansion opportunity provided to states. Other questions guided the team to analyze Minnesotas health care reform process in the early 1990s, including implementation of the Prepaid Medical Assistance Program (PMAP) for low-income persons. Background information on these activities and on demographic changes were factored into the analysis and were added to the report.
The first draft of the report was completed in February, 2001, and circulated to administrative staff at MDH and MDHFS. After further revisions the report was sent to every survey respondent, to the Minnesota Department of Human Services (DHS), and to the NHCN for review.
Phone and face-to-face follow-up discussions were conducted with survey respondentsas well as with DHS staffwho replied to the teams request for feedback, providing yet another wave of data.
The Minnesota Council of Health Plans hosted two meetings in the summer of 2001 with the plans representatives and the survey team. All plan representatives had received copies of the report in advance. In the first meeting, the survey team presented highlights of the report for discussion. The plans asked for additional time to respond. The second meeting, a month later provided the plans with the opportunity to respond to the survey team. At this point, the health plans added more detailed information about their care coordination activities. These additional data were added to the analysis and to the report.
Finally, it appeared that no further themes or data were forthcoming, and accordingly, the data-gathering/analysis phase ended.
ORGANIZATIONS INCLUDED IN SURVEY
Community Health Centers/Clinics
Cedar Riverside Peoples Center
Community University Health Care Center
Family Medical Center
Fremont Community Health Services (includes Fremont, Sheridan, and Central Avenue Clinics)
Hennepin County Womens and Childrens Health Program
Indian Health Board of Minneapolis
Model Cities Health Center
North End Health Center
Pilot City Health Center
Southside and Green Central Community Clinics
West Side Community Health Services
Public Health Nursing Agencies
Bloomington Division of Public Health
Minnesota Visiting Nurse Agency
St. Paul - Ramsey County Department of Public Health
Blue Cross/Blue Shield
Metropolitan Health Plan
U Care Minnesota
Abbott Northwestern Hospital
Fairview Southdale Hospital
Fairview University Medical Center
Hennepin County Medical Center and OB Clinic
Minneapolis and St. Paul Childrens Hospitals
St. Johns and St. Josephs Hospitals (HealthEast)
Page 8 of 21