Dr. Bell is president and chief executive officer of Community Mental Health Council, Inc., and professor of psychiatry and public health at the University of Illinois at Chicago, 8704 S. Constance Ave., Chicago, IL 60617 (e-mail: email@example.com). Dr. Sowers is director of the Center for Public Service Psychiatry at the Western Psychiatric Institute and Clinic, University of Pittsburgh. Dr. Thompson is associate director for medical affairs, Center for Mental Health Services, Substance Abuse and Mental Health Services Administration, Rockville, Maryland, and associate professor of psychiatry and public health, University of Pittsburgh.
In an effort to inform revisions of DSM-IV and ICD-10, New Zealand colleagues developed a survey of psychiatrists and psychiatrists-in-training regarding classification in psychiatry (1). Considering the plans to develop DSM-V (2), we thought it was important to understand the perspective of community psychiatrists regarding various aspects of DSM-IV. Accordingly, we surveyed the members of the American Association of Community Psychiatrists (AACP; www.comm.psych.pitt.edu) for their views on the general features of DSM-IV. AACP's membership consists of psychiatrists working in public and community settings.
Using Survey Monkey technology (www.surveymonkey.com), we converted the New Zealand survey questions to an online format after obtaining permission from our New Zealand colleagues. AACP's Webmaster e-mailed the survey to AACP's 600 psychiatrists with a cover letter requesting a response to the survey.
The survey instrument first asked for a single answer from 16 choices regarding the most important purpose of a diagnostic classificatory system. Questions also asked for a single answer about who should use a diagnostic classification and whether primary care practitioners should have a simpler system or use the same one that mental health specialists use. Respondents also chose their ideal number of diagnostic options (ten to 30, 31 to 100, or over 100).
Respondents rank-ordered, from 1 to 5, whether multiaxial systems should use the present DSM-IV axes, seek axes most useful in guiding treatment and care planning, or seek axes most relevant to understanding pathogenesis. In regard to future classification systems, practitioners were asked to rank-order two choices: should future systems develop classificatory systems designed to inform on cause or pathogenesis that are separate from those for clinical management decisions and care, and should they include a classificatory system, with many compromises, that conveys information concerning causes, prognosis, and treatment decisions? Respondents were also asked whether they routinely, sometimes, or never used axes 1, 2, 3, 4, and 5 of DSM-IV. Respondents were also asked for a single answer regarding whether DSM's present classifications are useful and reliable in practice regardless of service user ethnicity and culture, are sometimes difficult to apply across cultures, are too often unreliable or inappropriate where clinician and service user are of different cultures, or are overembedded in concepts or values derived from European culture. Finally, respondents listed their age and gender and chose the area or areas in which they currently (mainly) practiced clinically.
Survey Monkey is an online technology that allows users to develop multiple-choice items, rating scales, or open-ended text survey questions. It automatically tallies numerical responses and lists responses to open-ended questions. This research was anonymous, and because methods did not involve manipulating respondents' behavior and did not stress the respondents, the Community Mental Health Council's institutional review board reviewed the research through its expedited review procedure and approved it.
Of the 600 member psychiatrists requested to respond to the survey, 152 (25%) did so. Thirteen percent of the respondents (N=17) were under age 35, 34% (N=45) were 35—50, and 53% (N=71) were over 51 years of age. Two-thirds (N=88) were men, and one-third (N=45) were women. In regard to their clinical practice, 23% of respondents (N=31) focused on children and adolescents, 89% (N=117) focused on adults, 16% (N= 21) focused on forensics, 20% (N=27) focused on addictions, 21% (N=28) focused on older adults, and 9% (N= 12) focused on other areas of practice (totals sum to more than 100% because respondents could choose more than one focus).
See Table 1 for responses to the question, "From your perspective, which is the single most important purpose of a diagnostic classificatory system?"
Only 14% of AACP respondents (N=21) felt that diagnostic classification systems should be used solely by physicians or psychiatrists. Nearly half of the respondents (N=73, or 48%) were in favor of use of a diagnostic classification system by physicians, psychiatrists, psychiatric nurses, psychologists, any member of the multidisciplinary team, consumers, and others, with 24% (N=36) favoring use by any member of the multidisciplinary team (excluding consumers and others) and leaving 14% (N=21) who preferred to exclude consumers and nonphysician professionals (psychiatric nurses, psychologists, and multidisciplinary members) from use. In response to the statement "Primary care practitioners should have … ," 70% of respondents (N=106) checked "the same classification system as specialists in mental health service/clinicians," and 30% (N=45) checked "a modified/simpler classification system." Forty percent of respondents (N=59) would prefer 31 to 100 diagnostic options, 33% (N=49) would prefer ten to 30, and 26% (N=39) would prefer over 100; 3% (N=5) skipped this question.
The group was nearly evenly split (49% and 51%) in rank ordering whether future classification systems should "develop separate classificatory systems designed to inform on cause/pathogenesis from those for clinical management decisions and care" versus "include a classificatory system, with many compromises, which confers information concerning causes, prognosis, and treatment decisions." In regard to future DSM axes, respondents' first-ranked choice was "seeking axes most useful in guiding treatment/care planning" (N=81, or 61%), their second-ranked choice was "use the present DSM-IV axes" (N=53, or 43%), and the third-ranked choice was "seeking axes most relevant to understanding pathogenesis" (N=51, or 41%). Ninety-two percent of respondents (N=121) reported routinely using DSM-IV axis 1; 73% (N=95) reported using axis 2; 75% (N=98), axis 3; 55% (N=72), axis 4; and 48% (N=62), axis 5.
Regarding the question of cultural sensitivity, more than half (N=74, or 55%) of the AACP respondents noted that DSM-IV was difficult to apply across cultures, whereas 38% (N=51) noted that DSM-IV was useful and reliable in their practice regardless of ethnicity and culture. Nearly one-third (N=41, or 31%) felt that DSM-IV was overembedded in concepts or values derived from European culture, and 27% (N=36) responded that DSM-IV was "too often unreliable or inappropriate where clinician and service user were from different cultures."
Although a 25% response rate to an e-mail survey is a limitation of this survey, the median survey response rate for e-mail surveys is 26% (3). This median response rate represents a quarter of AACP membership, so we maintain that this information should be a consideration for crafting DSM-V.
Researchers have argued that DSM-IV criteria hinder investigation into the etiology, pathophysiology, and genetics of mental disorders and have proposed changes to make DSM-V more useful for research (4). Half of AACP respondents want a classification to be a reliable interclinician tool, in agreement with First and colleagues (5), who proposed that future revisions of DSM empirically demonstrate improvement in clinical utility. The inconsistency between researcher and AACP respondent perspectives may raise the question of whether it is time to begin thinking of a dual classification system, that is, creating one that is clinically useful and accessible to a variety of stakeholders and another that is a more highly differentiated version for the purposes of research.
There was significant support among AACP respondents for patients' participation in the diagnostic process. This is consistent with recent interest in and movement toward recovery-focused care because it promotes the idea that individuals must be well informed to make meaningful choices in their treatment (6,7,8). Consistent with this concept, AACP respondents indicated that a classification system should not be so arcane that only highly trained professionals can understand and use it. Further, despite 284 potential diagnoses in DSM-IV, only 26% of respondents would prefer having more than 100, and nearly 74% of those surveyed would prefer fewer than 100, indicating that AACP's respondents wish for a simpler diagnostic system. Finally, nearly a third recognized the "monocultural ethnocentrism" (9) in DSM-IV, and 25% noted DSM-IV is often unreliable or inappropriate when clinician and service user are from different cultures. This finding suggests that DSM-V should acknowledge and provide guidance on cultural differences (10).
Clearly, the members of the AACP who responded to this survey would prefer a DSM-V that is clinically functional, has fewer rather than more diagnoses, and is more culturally relevant than previous editions.