Archives   Calendar    Coordinating Center    Multisite Activities    
Project Description  COSP References  Study Sites    Search Our Site   
Table of Contents    Upcoming Meeting Agendas

______________________________________________

Steering Committee Notes: July 09,1999

July 9, 1999

Jean Campbell, moderator

Crystal Blyler
Sally Clay
Pat Corrigan
Dianne Cote
Jaime Delgado
Louetta Hix
Matt Johnsen
Anika Keens-Douglas
Ed Kramer
Sarah Lickey
Sandy Lint
Danyelle Mannix
Brian McCorkle
Betsy McDonel
Carol Mussey
Ruth Ralph
Miriam Righter
Jean Risman
Sally Rogers
Mark Salzer
Bonnie Schell
Steve Segal
Carol Silverman
Tom Summerfelt
Greg Teague
Daphne Walker-Thoth
Jim Ware, consultant
Brian Yates

Announcements

Jean Campbell announced that the Coordinating Center is finalizing arrangements for the upcoming Steering Committee face-to-face meeting. She asked that attention be paid to the deadlines. Anyone with items for inclusion in the meeting packet needs to submit them to the Coordinating Center by July 23^rd.

The Coordinating Center is in the process of testing the listservs as part of its reoganization of the telecommunications system.

Miriam Righter of the Iowa study site was welcomed to the call. She will be participating in future Steering Commmittee teleconferences and has joined the local Consumer Advisory Panel.

Jim indicated that the goal of the consultation was to determine how effectively the study sites could join together in the multi-site study. He likened the approach to the COSP multi-site study with that of the aging project in that a diverse group of sites was brought together under a broad GFA. The aging sites had different interventions, populations, and some other different elements. The ability to make fundamental changes to pursue multi-site goals is difficult in these instances, but the value is the opportunity to produce definitive and generalizeable research that is more powerful and influential.

He identified the following three elements of a multi-site study:

Eligibility criteria (there has to be agreement about criteria for enrollment of participants and it has to be generalizeable).

Randomization.

Commonality of intervention strategies (these strategies don’t have to be identical, but they have to have in common critical ingredients that result in a benefit).

If there are elements that are so fundamentally different among the sites, there may need to be a cluster approach around the different forms of intervention. There may need to be work toward the common protocol by grouping the different forms of interventions. Then a determination can be made whether or not the study can be taken to the next level.

In regard to inclusion/exclusion criteria, Jim stated that participants don’t have to be identical, but COSP needs to define the population of potential participants from whom one believes there is a shared common benefit. Participants who would benefit should not be mixed with those who would not. As a general proposition, the screening of participants is a good idea. First you want to demonstrate the effectiveness of COS and second think about generalizability. The result of enrolling participants who do not take advantage of the intervention is that the sample size requirement increases and the effect of intervention is diluted. The results can’t be generalized to those who don’t take advantage of COS.

He recommended the selective enrollment of participants. Jean ask about including people who have had experience in other types of programs that have similar effects to those of COS. Jim said if we believe that drop-in centers, peer support services, and other types of COS have elements in common that benefit participants regardless of the form of services provided, then those who receive these services in any form must be excluded. If we believe there are different populations that benefit from the different types of services, we would move toward several cluster services and then those with peer experience would not have to be excluded from drop-in centers and vice versa. He emphasized that we want people who are going to be responsive and benefit. Miriam Righter said the drop-in centers in Iowa are not consumer-operated and suggested the importance of our being clear about our definitions.

Jim added that enrolling participants who will not receive differential benefit from the two programs is counterproductive. In fact, it is worse than enrolling no one. To enroll someone who will not take advantage of the COS is worse than enrolling no one and to enroll someone already receiving services will hinder the ability to show differences and we won’t be able to differentiate. Matt Johnsen suggested that if we set things up in a way that data would be poolable at the end, then even if we have to stop at a point before we are able to pool across all sites, we have left open the option down the line. He argued for maintaining options. Jim said that in the aging multi-site they found that the sample size was not the main obstacle. Once they focused on the hypothesis, they found centers easily had the capacity to enroll the numbers needed. It may be that if COSP does power calculations and hypotheses, It might be able to set enrollment goals a little lower.

Tom Summerfelt said in Tennessee, they didn’t know how to predict who would benefit, so they adopted a broad-based approach to participant recruitment. Matt said at one point the idea of recruiting individuals and actually doing some screening to determine their receptivity to COS was mentioned. Sally Rogers said the drawback was that when people have enough information about the intervention, they may self-select.

Steve Segal asked: If we are measuring different interventions for different services populations, which is a more conservative approach? Jim said the possibility that this program would evolve into a three multi-site efforts with each focused on one intervention using a meta-analytic approach. Betsy stated that the idea of the multi-site hasn’t been given up, there is still some common ingredients work that needs to be done. A comprehensive list of common ingredients has been generated.

Jean Campbell stated that the multi-site calls for the most stringent exclusion criteria if we have a lot in common. We will need to exclude people who have been in any kind of COS if the common ingredients exist in any of those places. Sally Rogers said we haven’t studied common ingredients well enough to know how to use them for exclusion criteria. We don’t know if common ingredients are the things that cause change in people. We still have to think about aside from common ingredients, the unique services at the sites that might be active ingredients. Sally said exclusion/inclusion criteria has to be specified in the next few weeks and we won’t be far enough along with the common ingredients to use that information to develop these criteria. Jean Risman said we have some values in common in COS, but we don’t know if you look at what actually goes on how those values translate over into activities at the site level and if they translate the same way and if that translates into change and outcomes. Jean Risman said we have already put the cart before the horse. Betsy said the Common Ingredients Work Group can do translation of common ingredients and how they are manifested.

Betsy didn’t agree with Sally’s comment that we are not far enough along with the common ingredients. She suggested setting the criteria now. Mark Salzer agreed and said the assumption of the GFA is that there are common ingredients and we need very similar inclusion/exclusion criteria and we need to select people who have some interest in the program.

Jim said we could try to identify the subgroups of the population in which interventions are effective, but there is a high level of disaggregation there. Stephen Fischer asked about stratified random sampling. Jim said the sample would have to be too big.

Jim suggested identifying the population in which the intervention would most likely be effective and test the intervention. Then go to the next level of subgrouping to identify those who would benefit by gender, social class, type of presenting problem, etc. Betsy said that COSP did agree to narrow recruitment down to people who would benefit. The restriction of range would not permit for a strong secondary analysis of this type.

Mark said if we have a two-level analysis with the first being fully poolable across all sites and the second poolable across interventions, then we should go with the most self-selected sample of people who have demonstrated some interest in COS.

Sally Rogers said a second issue is participants’ prior involvement in COS. Stringent criteria could present problems for some sites. We will have vast differences in our study population since Philadelphia focuses on participants with substance abuse problems, Iowa has some unique ethnic differences, and California works with a homeless population. Jean Campbell agreed that the prior involvement issue is one with which COSP is still struggling. She said participation in peer mutual support groups was ubiquitious, but if we exclude this, we will destroy the study since so many people have this type of experience. We may not be able to exclude mutual support group participants, but we might be able to exclude people who have had experience with drop-in or educational programs.

Sally presented the following three scenarios that the Logistics Subcommittee group developed on exclusion criteria:

#1: If applicant has received any consumer-operated services similar to those of the COSP under study at the site in the past 6 months, that person will be excluded from the study. The name of the place where services were received would be obtained from the applicant and the study site would be responsible for determining if the place was a consumer-operated services or not using the following definition:

A COS is a program in which hiring/firing decisions are made by consumers, the program is managed by consumers, services are delivered by consumers, AND the program is not administratively responsible to a larger non-COS entity.

Under this scenario, the intensity of service received in other COS would have to be measured at baseline.

#2: If an applicant had more than three contacts with any consumer-operated service as defined in the GFA (drop-in, educational, or peer support) in the past 6 months, that person will be excluded from the study. This scenario would be implemented for three months and revisited to see if recruitment problems are resulting.

Under this scenario, intensity of involvement with consumer-operated services would be measured in the common protocol.

#3: This scenario is the same as #2 except it would not allow applicants to have had

any contact with consumer-operated services.

She said #3 would be the cleanest. Jim said the theshold of three contacts is a judgment about therpuetic impact. He said any of the scenarios would be defendable statistically. Jean Campbell said a fourth scenario would allow mutual support experience. Jim said since this is the largest population, to allow them we would have to demonstrate that COS adds value over and above benefits people are already receiving.

The first special call will be July 16 at 1 p.m. Eastern Time. Jim was invited to participate on the call.

Jaime Delgado reported that after review of the issues that emerged during the site piloting of the common protocol, it seemed that the issue was not the questions themselves, but with the way the questions were posed and the lack of adequate interviewer instructions. Jaime moved that the following recommendation be accepted:

Now I’m going to ask you some questions regarding discrimination that you may or may not have ever experienced within the last three years.

Interviewer note: The whole question could be handed to the respondent to fill out with the interviewer just reading it out loud for clarity, and to encourage timeliness.

(Hand respondent card)

1. I do not believe that I have been discriminated against......00

(Please skip to next section)

2. I believe that I have been discriminated against due to my:

(Please indicate all that apply)

Mental disability/ability 01

Race 02

Gender 03

Sexual Orientation 04

Religion 05

Country of Origin/Ethnicity 06

Age 07

Economic Circumstance 08

Physical Disability 09

Arrests with jail time 10

NANS........................ 88

NASK....................... 99

If yes, to any in Question 2:

3. This discrimination occurred in relationship to:

(Hand respondent time card)

Categories Yes No If Yes, How stressful for you*?   NANS NASK

Employment 1 0         1----2----3----4----5---6                       8     9

Education     1 0         1  ----2----3----4----5---6                      8     9

Housing         1 0        1----2----3----4----5---6                         8     9

Law 
Enforcement 1 0        1---2----3----4----5---6                         8     9

Pub.
 Accomm.     1 0         1----2----3----4----5---6                        8     9

COS               1 0         1----2----3----4----5---6                        8     9

TMHS           1 0         1----2----3----4----5---6                         8     9

Other Agencies

(Govern-
mental)         1 0         1----2----3----4----5---6                         8     9        

Other

Specify __________ 

                        1 0     1----2----3----4----5---6                            8     9

* Given time card

The motion was seconded by Jo-Anne O’Connor. Steve Segal spoke against the motion stating that this is an important issue, but in the common protocol we must focus on the intervention and what has the most influence on the outcomes we are seeking. There are more directly-related issues that need to be addressed in the protocol. He suggested that this issue be included at the site level. Miriam said one of the benefits of peer-operated services is the ability to deal with stigma and discrimination. The measure is measuring the amount of stress that is has caused the individual. If there is a reduction, this would constitute a positive outcome. Mark Salzer said it also relates to one of the the big six outcomes – social inclusion. He said his concern was with the scale. Jean Campbell added that in the GFA there is a whole section on how this initiative will be cultural competent – this represents our first effort that is objective. Steve pointed out that in the social inclusion measure there is a question about being treated differently because of mental status. Matt said he was very interested and the task force has come up with an instrument that will allow us to access experiences of discrimination and this is an important contribution.

The motion passed with 15 votes in favor and 4 votes against. Jean Campbell said this doesn’t mitigate the need to do some piloting of these questions.

On behalf of the Side Effects Work Group, Jean Risman made the following recommendation in the form of a motion. It was seconded by Ruth Ralph and passed unanimously with a friendly amendment by Sally Rogers indicating that we identify early on a provision for looking at the open-ended questions and formulating some coding for them.

The Subjective Side Effects Rating Scale (SSRS) be included in the common protocol. This would replace the side effects scale, adapted from the Hopkins Symptom Checklist, currently in the protocol.

The reason for including the a side effects measure is the recognition by the work group that side effects are themselves a potential outcome of treament and that side effects can effect the outcomes that we are measuring in this study. An international study by Ronald Kessler found that 45% of people internationally and 52% of people in the US stop treatment because of the side effects of psychiatric meds (poster presentation, American Psychiatric Association Conference, 1998). Panic attacks are common in people in treatment with long term neuroleptics (Higuchi et al, 1999) as is depression (Harrow et al, 1994). Sexual dysfunction is also a common side effect of psychiatric meds (Clayton & Shen, 1998) Side effects of psychiatric drugs have been shown to effect quality of life (Naber, 1995; Sullivan et al, 1992). Jean Cambell cited the Wellbeing project in hypothesizing that side effects effect social inclusion. Other outcomes that might be effected by side effects are recovery, empowerment and satisfaction. The SSRS was chosen as the instrument to measure side effects because it is an already extant measure with established psychometric properties. (I have contacted Peter Weiden, the developer of the scale, and am waiting to hear back from Peter on the psychometric data. Peter is a consultant for the Connecticutt project and would most likely be available to consult with the national Cosp concerning the SSRS. There are no copyright issue involved in using the scale since it was developed under a federal grant.).

The SSRS should be used only to measure the currently occuring side effects of psychiatric medication prescribed by doctors.

The introduction to the scale would be modified to reflect this. This would include medication taken orally over the past week and medication taken by injection which may not have been administered in the past week but would still be exerting an effect. Over the counter alternative treatments like kava kava or St John's wort woudl be excluded. Also excluded are psychiatric effects of drugs prescribed for non-psychiatric conditions. We realize that it will be impossible to tease this apart in people who are taking both kinds of drugs.

The logic model committee modify the current logic model to inlcude the potential of differential outcomes based on differing program models. This would include both intermediate outcomes and long term outcomes.

Some COS programs provide direct education on side effects while others don't address the issue directly but may provide a supportive community where side effects can be discussed and informal education and support is given. We need to take the possibility of differential effects by program type into account.

The original purpose of including a side effects measure was to try to tease apart the confounding of side effects and symptoms. While the work group continues to recognize the importance of this issue, it has decided not to address it within the common protocol. The reasons for this are two fold. First there is no extant measure which can be used to accomplish this and secondly and, perhaps more importantly, the issue of the confounding of symptoms and side effects does not fit the logic model and therefore the stated purpose of this study. We hope to be able to address this issue of confounding in the discussion section of the final report. There is ample evidence that the confounding exists and we believe that it is important that future research take up this issue directly including developing appropriate measures and statistical techniques to tease apart symptoms and side effects.

Steve Segal moved for adoption of the following questions for inclusion in the common protocol: Do you have a physical disability that seriously affects your ability to do your daily activities? The answer would be "yes" or "no". If the answer is "yes", the participant will be asked: What is your physical disability? The motion was seconded by Miriam Righter. Sally Rogers amendment that the interviewer be given a list of physical disabilities to do the coding and that attention be paid early on in piloting and coding of physical disabilities was accepted. The list will not be read to participants. Louetta Hix’s amendment that the following follow-up be added was also accepted: Do you have any other disability that seriously affects your ability to do your daily activities other than a psychiatric disability? If so, what? The amended motion passed unanimously.

Several Steering Committee members recommended that instead of holding off on the interviewer training until after the pilot that these two activities occur concurrently after the August face-to-face meeting. It is recognized that the training materials, including the videotape, will likely have to be amended. Procedures will probably continuously change. No site pilots will be conducted during the training. Jean Campbell will come up with some suggestions for implementation of this idea. Jean will also send letters out to principal investigators to find out who plans to participate in the pilot, how many participants they will interview, and whether or not IRB approval or exemption has been sought.

Jean volunteered to revised the common protocol time line and present it for discussion at the July 30^th teleconference. She asked that anyone who has any last recommendations on scale in the common protocol submit them to her in writing prior to the call.

July 16^th – special Steering Committee teleconference on selection criteria/randomization at 1 p.m. Eastern Time. July 30^th – regular Steering Committee teleconference at 3 p.m. Eastern Time.