Three major questions guided our research over the past several decades. First, we wanted to know if there were differences in family functioning between normal families and families in which one member has a psychiatric or nonpsychiatric, medical illness. Our clinical impression that there were differences needed to be examined systematically. Second, we wanted to test the assumption that a family’s functioning had a measurable impact on a patient’s course of illness. Third, we wanted to test the PCSTF, the family treatment approach that was based on the MMFF.50 Before addressing these questions, we needed to develop a set of constructs that were reliable, valid, clinically meaningful, and empirically testable. Epstein and his colleagues embarked on a series of studies described in previous chapters. As we noted, clinicians and researchers worked together in an iterative process, testing, questioning, and refining concepts until satisfied with a prototype of a family functioning model, a treatment approach, and research assessments. These prototypes were used to address part of the first question listed above, that is, family functioning in normal families. Readers can appreciate the complexity of the issues involved when a series of studies was needed just to address what constitutes normality.1,7,51,52

Once we built the framework for a valid and reliable model, and instruments to assess the model, we were free to examine family functioning of families across a wide spectrum and in a variety of contexts. We compared family functioning in families in which a member had a psychiatric illness, a medical illness, or both. We analyzed the family’s functioning in families

with a member who had a specific illness (e.g., major depression, rheumatoid arthritis) or a specific clinical characteristic such as suicidality or stroke. We conducted both short-and long-term follow-up studies and examined the relationship between family functioning and course of illness. We examined different stages of the illness (acute, 6-, 12-, and 60-month follow-up) to look at questions that were cross-sectional in nature and we used several time points to analyze longitudinal issues. Our studies included families with a member hospitalized for an illness, family members treated as outpatients, and community controls. We compared family functioning across different cultural groups. Finally, we analyzed data by individual family member characteristics such as role (patient, parent, caregiver), gender, and marital status.