Commotion in Psychiatry, Part III: Deinstitutionalization, Rationalization, Medicalization, and Fuzzy Decision-Making
A third motive for the switch to a criterion-based model of diagnosis (instead of a prototype-based approach) can be found in multiple changes in US society. In the 1960s and 1970s, mental health became an important topic for diverse stakeholders. Overall it was observed that during these years the need for mental health care increased dramatically.
Between 1955 and 1971 the number of patients applying for treatment doubled: in 1955 an average of slightly over 1 in 100 of all people had had a treatment episode, while in 1971 the ratio was almost 2 in 100 (Strand 2011). Despite strong investment in community mental health care and in deinstitutionalization (Grob 1987), more patients than ever before stayed in psychiatric institutions. This influx of patients meant that traditional long-term treatments could no longer be offered to everybody, which inevitably gave rise to public discontent with regard to common treatment methods. Gradually, psychotropic medication became the treatment of choice (Strand 2011; Decker 2013). In the context of the pressure put on the mental health care system, checklist-based thinking was promising. It assured a break with tradition, and it offered a transparent model for characterizing mental illness, which could then be applied to the masses in a relatively short time span.
While third parties, like the pharmaceutical industry and private and public insurers, didn’t actively contribute to the creation of DSM-III, they certainly welcomed the system with open arms, as it capitalized on many of their interests and concerns (Strand 2011). With the overall increase in the number of individuals applying for treatment, financial pressures increased and psychiatric care became an activity that administrators wanted to manage following the management model of industrial companies. Insurance agencies, suspicious of long-term psychotherapeutic treatments, were increasingly interested in a rationalization of mental health care. For them a straightforward criterion-based classification system, which would also facilitate research on more effective treatments, was most promising.
The pharmaceutical industry in its turn, which by then fully endorsed the practice of randomized controlled trials, approved the criterion-based approach because of its use of explicit inclusion and exclusion criteria (Healy 2008; Strand 2011). This allowed them to delineate specific disorder groups that could be linked to specific pharmaceuticals. Such an approach was promising, not only because it fuelled the hope that pharmacological interventions could be linked with specific disorders, but above all because it enabled the marketing of specific products for specific target groups. By 2004 the industry was spending approximately twice as much per year on marketing (US$57.5 billion) than on research (US$31.5 billion) (Gagnon and Lexchin 2008).
Indeed, as Strand (2011) indicates, this exploitation of the DSM by the pharmaceutical industry, which can certainly be observed nowadays, only took off following the publication of DSM-III. An unfortunate side effect of the DSM-III is that it gave rise to disease mongering in psychiatry, that is, the redefinition of common discontent in psychiatric terminology: “Pharmaceutical companies are actively involved in sponsoring the definition of diseases and promoting them to both prescribers and consumers. The social construction of illness is being replaced by the corporate construction of disease” (Moynihan et al. 2002). Disease mon- gering has been documented for several conditions that are included in the DSM, including shyness/social phobia (Lane 2008; Moynihan et al. 2002); discontent/depression (Dehue 2008; Dowrick 2009); and inattentive behavior/ADHD (Philips 2006; Timimi and Leo 2009). Disease mongering not only generates medical consumption, but also undermines the validity of diagnostic categories: if both common misery and severe mental distress are lumped under the same label, the precise meaning of the diagnostic category ultimately becomes even more ambiguous.
While in the past three decades the pharmaceutical industry appears to have had special interest in the DSM, Allen Frances (2013), the chair of the task force that designed DSM-IV, indicates that the industry has had no direct influence on the panel for the DSM-IV. The author claims that while financial conflicts of interest have not compromised their decisions, the carefully considered decisions were certainly later exploited: “The drug industry played no role whatever in how DSM-IV was written, but played a deciding role in how it was misused” (Frances 2013, p. 76). Indeed, he suggests that individual experts’ fads and fancies perhaps informed decisions that were made, but he does not connect these personal whims to any kind of influence made by the pharmaceutical industry. It could nevertheless be argued that there was a clear impact at that level. The pharmaceutical industry and academics studying pharmaceuticals frequently collaborate. This no doubt gives rise to loyalty, which influences decision-making. For example, Lisa Cosgrove et al. (2006) studied “conflicts of interest” in panel members of the DSM-IV and concluded that out of 170 associates, 56 % had important financial connections with the pharmaceutical industry, among which were those focusing on “mood disorders” and “schizophrenia and other psychotic disorders,” which comprise the most frequently used diagnostic categories. In the panel working on the DSM-5 these conflicts of interest were even stronger: overall, 69 % of the 141 members had important financial connections with the pharmaceutical industry, reflecting a relative increase of 21 % compared to the DSM-IV panel members with such ties (Cosgrove and Krimsky 2012). More specifically, financial conflicts of interest of DSM panel members responsible for some new diagnoses in the DSM-5 (like the elimination of the bereavement exclusion criterion in major depressive disorder, binge eating disorder, and disruptive mood dysregulation disorder) and principal investigators of clinical trials for related drug treatments could be identified (Cosgrove et al. 2014). These financial ties certainly don’t prove that panel members were biased. Yet considering these sponsoring connections, it is not groundless to harbor at least a degree of skepticism about the scientific independence of DSM panel members.
If task forces working on diagnostic systems want to avoid discussions about the influence of the pharmaceutical industry, they simply have to be more careful in their selection of board members, and follow a policy that assures their independence. Moreover, they should be open about the decision-making processes that lead to endorsing or changing diagnostic categories. If commercial interests do not influence the choices made by panel members, minutes of panel meetings, and open communication by diverse panel members on the decision-making processes, would offer evidence of their neutrality. Such transparency was lacking more than ever before in the development of DSM-5. In fact, panel members had to sign agreements in which they guaranteed secrecy about all meetings. This goes against the kind of transparency that is crucial to scientific practice (Peerenboom 2002), something that the former DSM-III chair Robert Spitzer too believed: “When I first heard about this agreement, I just went bonkers. Transparency is necessary if the document is to have credibility, and, in time, you’re going to have people complaining all over the place that they didn’t have the opportunity to challenge anything” (Spitzer, in Carey 2008). If independent peers cannot assess the methods and arguments used in obtaining a result, the truthfulness of the final result is not only suspect, it is simply insupportable.
Contra this criticism it could be argued that between 2010 and 2012 three drafts of the DSM-5 were presented online for the general public, and that the task force requested feedback each time. In other words, by setting up an online feedback platform, for the first time in the history of the DSM, the DSM-5 permitted wide-ranging input from the general public throughout its development. The DSM-5 (p. 8) states, “These three iterations of external review produced more than 13,000 individually signed comments on the website that were received and reviewed by the work groups.” However, this promising development was somewhat undermined by the fact that it was (and remains) unclear as to how these comments were reviewed or analyzed, and, moreover, whether they were considered at all during the process of internal discussions and final decision-making. The same is true for commentaries endorsed by “thousands of organized petition signers” (DSM-5, p. 8). As preliminary versions of the DSM-5 were presented online, several professional societies, as well as patient and advocacy groups, submitted widely signed petitions. For example, the British Psychological Society publicly criticized the inclusion of two new diagnostic categories in the preliminary version of DSM-5: disruptive mood dysregulation disorder and attenuated psychosis syndrome. Both categories were based on vague criteria and had the potential to promulgate waves of false-positive diagnoses. The vice-president of the DSM-5 task force (Darrel Regier in The Psychologist 2011) responded to this criticism with a message that was simultaneously polemical and vague, using imputation (e.g., pointing to the British Psychological Society’s reference to Joanna Moncrief, who endorsed ideas of Thomas Szasz) but no evidence or precise arguments that could substantiate the new categories. In the end, disruptive mood dysregulation disorder was retained as a new disorder category in the DSM-5; attenuated psychosis syndrome was cast as a disorder category for which future research is needed. Aside from the question as to whether including these diagnostic categories is acceptable, it is certainly unacceptable that the so-called evidence-based method to which decisions in DSM committees and workgroups adhere is not openly available to the scientific community.