An Application of the Use of Outcomes Measurement with Multiply Diagnosed Low Income Ethically Diverse Clients
Rosemary Moynihan, L.C.S.W.
Marjorie H. Royle, , Ph.D.
St. Joseph's Hospital and Medical Center
Many social scientific research methods were developed on the college sophomore, who traditionally has been young, white, middle-class, well-educated, generally motivated, in good health, and a captive audience. This paper discusses how such methods must be adapted for people who may be none of the above, those in the inner city who are infected with HIV. Examples are taken from our experience at St. Joseph's Hospital and Medical Center evaluating a Ryan White Special Project of National Significance that provides mental health services to such a population in a manner that is integrated with their primary medical care.
Trade-offs must be made between standard scientific methods and the ability to access the population of interest. Traditional measures must be modified. Surveys must be shortened and simplified, with decreased reading levels, and may need to be translated into Spanish and other languages. To do so, precision of measurement, obtained by using multiple items to measure a construct, may have to be sacrificed to get any measure at all without eliminating many of the people in the target group. Respondents, particularly those who are unfamiliar with surveys or have limited literacy, may need to be assisted by having a research assistant read the survey instruments and explain some questions, even at the risk of introducing bias.
Thought must be given to motivating people to participate. Compensation in the form of money, gift certificates, or other meaningful incentives is important, as are good outreach and interpersonal skills on the part of these enrolling participants are a captive audience, such as a waiting room, is also important.
Alternatives to paper-and-pencil tests should also be used. Individual or group interviews may be more appropriate and provide data that is richer, more nuanced and more meaningful in terms of the issues important to the respondents than surveys developed by researchers from outside the particular setting. Archival measures, often already collected by the organization, such as appointment-keeping, numbers of referrals, and use of other services, can provide additional important information.
Finally, in clinical settings, traditional experimental models with randomized control groups raise many ethical and implementation issues. Quasi-experimental designs, and designs that use a variety of methods to attempt to balance or cancel out the problems and biases inherent in each are alternatives which should be explored.
St. Joseph's Hospital and Medical Center is part of a consortium with Albert Einstein College of Medicine and Montefiore Medical Center to evaluate the effectiveness of providing mental health services for people with HIV in a manner that is integrated with primary medical care. Both St. Joseph's and Montefiore were recipients of the original SPNS Mental Health grants. With a continuation grant, we are comparing our integrated site with the HIV program at the Barnert Hospital in Paterson, where mental health services are provided by referral to the Barnert clinic, rather than on-site, and Montefiore is comparing its clinics with mental health services on-site to its clinics that provide such services by referral. Michael Mulvihill, DPH, in the Department of Family Medicine at Albert Einstein College of Medicine, is the principal investigator for the research, although all three agencies have been involved in the design and the day-to-day supervision of staff.
At St. Joseph's, in the seven months since data collection began, we have enrolled 69 newly-diagnosed clients in the study, and at the Barnert, we have enrolled 44, numbers which are on target for the research plan. Most of our clients are African-American or Latino. Our research assistant has few refusals. The numbers enrolled are considerably higher than those at the Montefiore sites, although some of the difference is due to differences in the way the medical services are delivered (we have dedicated HIV clinics, and they do not, so enrollment efforts can be targeted). In the design of the study, considerable effort was put into developing a plan that was implementable in our setting, and we think we have succeeded.
Return to Program
Return to Abstracts