The Kaiser Family Foundation: Report - Is There A Common Ground?

Appendix B: Methods

I. Trend Analysis Methods

To profile recent affiliations involving Catholic hospitals/health systems, a database of 131 such transactions occurring between 1990 and 1996 was constructed in three steps. The database includes: (1) transactions involving community hospitals that affiliated with other hospitals or health systems, and (2) transactions involving significant organizational changes. The database does not reflect major corporate affiliation activity, transactions that were reported but later dissolved, or attempted affiliations that failed.

The first step in constructing the database was obtaining annual lists of hospital affiliations from two sources: the American Hospital Association (AHA) and Modern Healthcare, a weekly publication of Crain Communications Inc. that tracks health care industry news and trends. The Constituency for Health Systems at the AHA provided lists compiled between 1990 and 1993. Annual reports compiled by Modern Healthcare staff provided information about formal affiliations occurring between 1994 and 1996. Partnerships involving significant changes in control or financial statements of a hospital were listed, including: mergers, acquisitions, consolidations, joint ventures, and lease agreements. Looser forms of affiliation, such as management contracts or cases where a hospital joins a network, were not listed. For each transaction, the lists identified the names of the partners and the new organizations formed (if any), affiliation type, state, and year of affiliation.

The second step was identifying transactions involving Catholic parties. This was done by matching the above lists with the 1995 Official Catholic Directory, the recognized authority for publishing the list of "agencies, extensions, and affiliates of the church" (Showalter and Miles 1988). Since it is possible that a Catholic hospital that dissolved following an affiliation would not be listed in the 1995 Directory, the lists were also matched with the AHA guides to the health care field to identify hospitals belonging to a Catholic system. This strategy minimizes the possibility that a Catholic hospital affiliation is not identified for failure to determine Catholic identity in cases where the Catholic partner dissolved following the transaction.

The third step was to determine the type of ownership of the non-Catholic partner(s). Transactions involving Catholic partners were matched with the AHA guides for the appropriate years, and the ownership type of the non-Catholic partner(s) was identified as public, private not-for-profit, or for-profit. (For two transactions, the systems affiliating with the Catholic hospital were not listed in the AHA guides but were identified in the Directory of the Federation of American Health Systems, the trade association of for-profit systems.)

Using this information, the 1990-1996 database of Catholic hospital/health system affiliations was divided into three categories: (1) transactions involving Catholic partners only; (2) transactions involving Catholic and non-Catholic not-for-profit providers (including private not-for-profit and public organizations); and (3) transactions involving Catholic providers and non-Catholic for-profit organizations.

The limitations of the approach used to construct the database are as follows:

1. Affiliations that were not publicly announced or reported were not included in the source lists. It is not possible to locate such cases or to estimate their number.
   
   
2. Despite the fact that the source lists were matched against the Official Catholic Directory and the AHA guides, it is possible that certain transactions were not identified as involving a Catholic partner for failure to confirm Catholic identity. If this occurred, the number of such cases would be very small, however.
   
   
3. AHA and Modern Healthcare do not systematically monitor affiliations prospectively. It is therefore possible that the database includes affiliations that were announced but later failed.

II. Case Study Methods

A. Case Study Selection

Based on the database described above and information obtained from key informants, we selected twelve cases of affiliations between Catholic and non-Catholic providers (the most prevalent category of affiliations) as possible case study sites. The selected cases met the following eligibility criteria:

1. Cases were mixed with regard to affiliation type and included six mergers, two acquisitions, two joint ventures, one consolidation, and one long-term lease agreement.
   
   
2. Cases were geographically dispersed and represented all regions of the country.
   
   
3. Cases were mixed with regard to ownership of the non-Catholic partner: eight of the non-Catholic partners were private not-for-profit; three were public; and one was for-profit.
   
   
4. All but one affiliation occurred between 1994 and 1996. This reflects the fact that relatively few partnerships were reported between 1990 and 1993.
   
   
5. Cases were mixed with regard to the extent to which a Catholic presence was maintained post-affiliation. For example, the merger cases included some where the newly formed system was Catholic and some where the newly formed system was non-sectarian.
   
   
6. Cases varied in terms of the outcomes with regard to reproductive health services, as far as this could be ascertained from available information.

Of the twelve cases contacted and invited to participate, four agreed to participate in the study. Declining cases cited three main reasons:

1. a memorandum from the President and CEO of the Catholic Health Association (CHA), alerting members to the CHA Executive Committee's position that "there are serious negative consequences of this study for the whole health ministry" and recommending "that CHA not support nor participate in the study;"
   
   
2. fear that participation might expose the partners to criticisms around sensitive issues, including the provision of reproductive health services; and
   
   
3. busy schedules of hospital executives that did not allow time for hosting a two-day site visit.

The CHA memo, dated September 5, 1996, was distributed following numerous telephone conversations and a meeting between the project investigators and CHA representatives, in which the investigators explained the aims and methods of the project.

The characteristics of the four participating cases reflect all but one of the selection criteria noted above. All occurred between 1994 and 1996, they were geographically dispersed, they illustrated different affiliation types (including a merger, acquisition, consolidation, and joint venture), they had varying outcomes with regard to reproductive health services, and the non-Catholic partners were both private not-for-profit and public providers. The cases do not, however, include an instance where the Catholic partner is relatively dominant, such as acquisitions of non-Catholic assets by a Catholic institution or a merger of a Catholic and non-Catholic hospital into a Catholic entity. CHA's memo could have influenced the decision of such cases to decline participation.

It is possible that participating cases also differ from declining cases in two other ways. First, participating cases may represent relatively less conservative religious communities with fewer concerns about possible scandal. Second, the administrators in participating cases may be more comfortable with the progress of the post-affiliation process and therefore more willing to discuss their situations.

In any event, the four cases are not meant to be representative of all affiliations between Catholic and non-Catholic health care organizations. Rather, they are illustrative models of successfully negotiated affiliations and of outcomes for reproductive health services.

B. Case Study Methods

The chief executive officer at each of the contacted sites was mailed an introductory letter explaining the purpose of the project and the provisions for confidentiality. A site visit was conducted at each of the four sites that agreed to participate.

Each case study was conducted over a 1-2 day period during which a three-person team of investigators traveled to the site and conducted semi-structured interviews with key informants identified prior to the visit. At each site, an effort was made to interview senior administrators (e.g. the CEO, chief operating officer, vice-presidents, planning/marketing director), managers of women's services, members of the governing board, clinicians (including physicians and non-physicians), and ethicists. We were particularly interested in interviewing individuals who had participated in the affiliation process and continued to assume responsibilities within the organization. Interviews focused on the following topics:

1. history of the partners and community characteristics;
   
   
2. the affiliation process, including motivating factors, earlier attempts at affiliation (if any), key issues in operationalizing the partnership, post-affiliation governance, and current challenges;
   
   
3. status of reproductive services, including the range of services provided pre- and post-affiliation and decision elements in service provision; and
   
   
4. community impact of the affiliation. Each site visit also included a review of hospital and affiliation documents (e.g. organizational charts, annual reports, affiliation agreement documentation, hospital publications communicating the affiliation internally to staff and externally to the community, media coverage of the affiliation, proposals/plans for women's services) and of information on other community providers of reproductive health services.
   
   A case study report was prepared for each case, following a uniform format. Each report was sent to the CEO to be reviewed for accuracy, and minor revisions were then made. (See Appendix C.) The case study protocol, including the interview questions, is available from the investigators.