Guidelines for Customizing the SCID for Particular Studies
1. What do you really need?
Interviewers should not be using a SCID that includes diagnoses that they will not be assessing. A customized, stripped-down SCID not only saves trees, but causes less confusion for neophyte interviewers.
The SCID was designed with separate modules for each class of diagnoses. If there are classes of diagnoses that will not be assessed in a particular study, the entire module (and the corresponding section of the score sheet) may be eliminated. For example, in a study in which only Mood, Psychotic and Substance Use disorders are being assessed, SCID modules F,G, H and I may be eliminated. Note that since the assessment of Mood Disorders requires an inquiry about psychotic symptoms (Modules B and C, or B/C), and the differential diagnosis of Psychotic Disorders also involves assessing mood episodes, in most cases Modules A, B (or B/C) and D must be used together.
The elimination of entire modules is the simplest modification of the SCID. For many studies, modifications involve the elimination of specific disorders or subtypes, or time frames. For example, an investigator may need to document a current Major Depressive episode, but may not be interested in lifetime episodes or whether the Major Depressive Episode conforms to a particular subtypes (e.g. Post Partum, Melancholic, Atypical). In such a case, the SCID can be customized, eliminating the subtype or disorder pages from the module.
WARNING: Eliminating the subtypes often means changing the "GO TO" instructions on preceding pages, so these must be checked carefully when such a modification is made. Using the example above, at the bottom of page A5 (Item A28), after a Major Depressive Episode is documented, the instruction "GO TO NEXT MODULE" should be inserted so the interviewer does not continue with an assessment of a past Major Depressive Episode. The subtypes eliminated from the SCID should also be removed from the score sheet.
2. Exclusion Criteria
Treatment studies usually have specified exclusion criteria, so that if particular diagnoses are made, the subject is excluded from the study, and there is no point in continuing with the SCID interview. At each point in the SCID where an exclusion diagnosis is made, the instruction "EXCLUDE FROM STUDY - END OF SCID" should be inserted. For example, in a trial of a new drug for Major Depressive Disorder, the investigator may want to exclude all subjects with a history of a Manic Episode. If Item A99 (Manic Episode criteria A,B, C and D are coded "3") is true, the subject has a diagnosis of a current Manic Episode, and the EXCLUDE FROM STUDY statement should be inserted. The same is true for item A142, past Manic Episode. When the EXCLUDE FROM STUDY instruction is encountered, the interviewer should stop the SCID and gently dismiss the subject, perhaps offering other, non-study, options for treatment.
In many studies, the exclusion criteria do not overlap exactly with SCID conventions. In the example above the investigators may accept subjects with a history of drug or alcohol diagnoses, unless they are present during the past year. The substance abuse and dependence questions would have to be modified to inquire only about the past year, rather than lifetime. If a subject meets criteria for any substance abuse or dependence in the past year, the EXCLUDE FROM STUDY message should appear (e.g. items E6, E92, E93, etc.) In a slightly more complicated version of this assessment, the investigator may wish to inquire about a lifetime history of substance abuse or dependence, but exclude only those subjects who met criteria within the past year. It would then be necessary to go through all of the criteria, making the decision to EXCLUDE FROM STUDY only at the end of the Alcohol or Drug sections (e.g. items E21, E116, E117), after changing the definition of Current to correspond to the study definition (e.g. past year rather than past month).
3. Inclusion Criteria
A more complicated modification may be required for a study in which the goal is to screen in diagnoses during a specified period of time rather than for the past month, or for lifetime. For example, a drug study may have an inclusion criterion that requires a Major Depressive Episode with symptoms present during the past two weeks. The initial SCID questions for current Major Depressive Episode would have to be modified to inquire about the past two weeks rather than the past month. This change would also be reflected on the score sheet.
The foregoing are examples of the most common ways in which the SCID might be modified. There are many other possibilities, and we suggest that an investigator who is considering a modification should consult with someone at SCID Central about how to customize the SCID. Brief telephone consultation is free, but there will be a charge for investigators who need more extensive help in making major modifications.
4. Adding on to the SCID
Investigators may want to incorporate the collection of additional information into the SCID interview. For example, family history data that is to be coded may be collected during the overview with the addition of a few questions and items.
Rating scales, such as the Hamilton Depression Rating Scale, are sometimes inserted into the SCID. This process is more complicated than it may appear. It is necessary to pay attention to the time period being covered (usually past week for the Hamilton, and past month for the SCID depression criteria), and to make sure that the precise items in the scale, rather than the generally broader DSM-IV criteria, are being assessed. Investigators may want to discuss with a SCID expert the pros and cons of inserting a scale into the SCID or simply adding it to the end.....or the beginning....of a SCID interview.
5. Identifying your customized SCID
Anytime a SCID is modified, it should be identified as such on the cover page....e.g. "Modified for Use in the NIA Study of Late Life Depression, 9/95". In addition, a footer on each page should include this identification. The footer should also include consecutively numbered pages because it is easy to lose pages in the photocopying process, and to not realize they are missing until weeks, or months, later. Continuing with the example above, each page in the modified SCID should have a line across the bottom reading: NIA Late Life Depression - page 1....etc.
Another reason for including an identifying footer on each page is that students, teachers or other investigators will sometime pick up part of a SCID to reproduce, and it is important to identify that the severed SCID body part is not part of the standard SCID, and where it came from.
6. How much can you eliminate and still call it a SCID?
What differentiates the SCID from other instruments is a format that includes questions, diagnostic criteria and ratings. A very much slimmed down SCID that still has those characteristics is probably eligible for membership in the SCID family. For example, a study in which subjects are carefully pre-screened for current Major Depressive Disorder may need to confirm the diagnosis with only the A and B/C and D modules, and all the subtypes eliminated. In another study, investigators may be interested only in assessing PTSD. If the cover page identifies nature of the modification (e.g. SCID-PTSD), we see no reason not to acknowledge it as a member of the family.