A review of programs in Pennsylvania and Philadelphia for participation in the study began in 1994. Our interest in Philadelphia County was sparked by the fact that it had one of the few intensive family reunification programs in the country. However, at that time, the Philadelphia reunification programs were relatively new and were not serving enough cases to support an experiment. Discussions with a Social Work Administrator in charge of FPS in Philadelphia County, a strong supporter of rigorous evaluation of FPS, shifted our focus to consider the placement diversion programs that serve cases with serious substance abuse problems. Further discussions with the administrator and her staff took place in June 1995. Philadelphia was an interesting site for the study for three main reasons: 1) the FPS program was not a Homebuilders model like the other study sites, but instead focused on broader use of FPS including servicing intermediate risk cases and providing extended services to families; 2) Philadelphia's SCOH provided the opportunity to study differences in service intensity between the FPS and SCOH cases; and 3) the FPS program also provided a targeted look at families with drug and alcohol problems.
Support for the study was obtained in a series of meetings with DHS administrators in 1995 and 1996. The study was approved by the Philadelphia DHS Commissioner and the Pennsylvania DPW in the Spring of 1996. Many questions about implementation of the study arose in discussions with DHS middle managers. Central issues included eligibility criteria, random assignment, and case flow. These issues were resolved in meetings that occurred in the fall of 1996 and early 1997, under the leadership of the DHS FPS administrator and a FPS unit supervisor. Initial plans to obtain referrals for the study from the CRU were abandoned; instead it was determined that referrals would come directly from Intake.
The selection of programs included in the evaluation in Philadelphia was purposive. Programs that served families who were not referred through the CPS/GPS system were eliminated from consideration. The focus was on programs that specialized in cases with substance abuse problems. The study included two private agencies in Philadelphia that provided FP services, Abraxas and Tabor Children's Services, both of which also offered SCOH. A third SCOH agency, Youth Service, Inc., was included to insure that there were enough SCOH for cases that were randomly assigned to the control group. Thus, there were two FPS programs and three SCOH programs in the study. All five programs served the entire county.
Cases were enrolled in the study as follows. A DHS FPS unit supervisor reviewed all cases before they were referred for services to determine whether the case was eligible for the study. If eligible, the FP supervisor then determined whether there were openings in at least one of the FPS programs, and in one of the SCOH programs participating in the study. If openings were available, the FPS supervisor called Westat, where the case was randomly assigned by computer to either family preservation or SCOH. The case assignment was relayed immediately to the FPS supervisor over the phone. If one or both of the participating study agencies could not provide services at the time, the case was referred for SCOH in one of the agencies that was not participating in the experiment.