This project will use improved methods of longitudinal analysis to evaluate the multidimensional course of quality of life (QOL) in the four years following bone marrow transplantation (BMT), providing better descriptions of how QOL changes following BMT, thereby allowing patients to make more informed choices among competing treatments. The increased precision and accuracy offered by these methods may enable clinicians to intervene prospectively to minimize complications and smooth the path to recovery. The new methods for longitudinal evaluation are general, and will extend to other cancer populations. Building on an existing survey of 125 extreme (6 18 years) long term survivors and on extensive pilot work with new one year survivors, approximately 150 new patients will be recruited each year (600 total) for QOL assessments at 6 month intervals over four years, yielding a much larger longitudinal BMT sample than any previously studied. The questionnaire battery assesses functioning in the physical, role, emotional, cognitive, and social domains, and includes general psychometric assessments of affective and health perceptions. In addition, the battery includes BMT specific inventories, developed by project personnel, assessing late medical complications and demands imposed by recovery. Participating patients will complete all questionnaires in their residences, receiving and returning the packets by mail. The new longitudinal data collected for this project will enhance psychometric development by a (a) developing a briefer form of existing instruments amenable to more frequent administration, and (b) using structural equation modeling techniques to establish and improve the sensitivity to change of key indicators in the battery. Because of the large projected sample, standard longitudinal analyses will precisely describe the expected quality of life for patients conditional on survival to each assessment. Surviving subsamples, however, may selectively overestimate expected trends for the general population of new patients; thus a major focus of this study is the development of unconditional estimates of QOL trends that are not subject to this attrition bias. To obtain unconditional estimates of quality of life trends, the authors will a) employ mixed effects models as discussed by Donaldson; and b) develop endpoints, formally similar to Q TWIST, that use psychometric norms to weight survival times for QOL experienced during BMT recovery.
