Substantive changes to diagnosis of substance use disorders

The most substantial change recommended for substance use disorders is to eliminate the ‘abuse’ versus ‘dependence’ dichotomy. In DSM-IV, abuse is diagnosed when an individual endorses one or more of four criteria related to adverse social and interpersonal consequence of heavy substance use. A diagnosis of dependence is rendered for meeting three or more of seven criteria, which assess physiological and other consequences of heavy alcohol or drug use. The division between abuse and dependence was premised on the concept that abuse was a milder form of dependence. However, analyses from over 30 different studies, consisting of over 200 000 persons, revealed that 10 of the 11 DSM-IV substance use criteria form a unidimensional structure (e.g. Hasin et al, ^{Reference Hasin, Fenton, Beseler, Park and Wall1} Saha et al ^{Reference Saha, Compton, Chou, Smith, Ruan and Huang2} ). The four abuse criteria are not associated with less severe forms of the disorder, and instead span across the severity spectrum.

The only criterion that did not fit with the others and added little diagnostic information was that related to committing illegal acts. ^{Reference Hasin, Fenton, Beseler, Park and Wall1-Reference Agrawal, Heath and Lynskey3} This criterion has low prevalence rates, varies depending on the cultural context, and it is rarely endorsed in the absence of other criteria. Thus, a recommendation is to remove the illegal acts criterion in DSM-5.

One criterion is being recommended for inclusion. Strong desires to use or craving is a dependence criterion in ICD-10. The DSM-5 Substance-Related Disorders Work Group evaluated a number of cross-cultural epidemiological and clinical data-sets in which craving was assessed along with the other substance use disorder criteria. Craving fit with the other criteria and did not impact severity, discrimination or factor loadings. ^{Reference Hasin, Fenton, Beseler, Park and Wall1,Reference Agrawal, Heath and Lynskey3} For consistency with ICD-10 and because of support for inclusion of this criterion among many clinicians, craving may be included in the diagnosis of substance use disorders in DSM-5.

Thus, 11 criteria are proposed for substance use diagnoses in DSM-5, requiring consideration of the optimal threshold for diagnosis. Although abuse was diagnosed by meeting just one of four criteria in DSM-IV, a single criterion is not a syndrome. The DSM-5 Substance-Related Disorders Work Group evaluated data from epidemiological and clinical samples to determine agreement with DSM-IV diagnoses across thresholds ranging from ≥2 to ≥4 criteria. Because agreement was very good to excellent using a threshold of ≥2 criteria, it was selected. Although some have argued that a threshold of ≥2 criteria is too inclusive, this concern needs to be balanced with the need to diagnose – and treat – milder cases such as patients presenting in primary care settings who have underlying substance misuse problems. Severity indicators are proposed, such that meeting only two criteria represents a milder form of the disorder.

Substantive changes proposed for pathological gambling

In addition to changes in the diagnosis of substance use disorders, the chapter is likely to be broadened to include behavioural (or non-substance) addictions, such as pathological gambling. In DSM-IV and earlier editions, pathological gambling was classified in the Impulse-Control Disorders Not Elsewhere Classified section. Consistently high rates of comorbidity between substance use disorders and pathological gambling, genetic and physiological overlap, and similarities in presentation of some symptoms and treatment approaches led to the decision to include gambling alongside substance use disorders in DSM-5. ^{Reference Hodgins, Stea and Grant4-Reference Petry, Weinstock, Morasco and Ledgerwood8} With this move, the name of the disorder will also be altered to ‘gambling disorder’. Other terms such as ‘compulsive gambling’ and ‘problem gambling’ were considered. However, the term ‘compulsive’ is commonly applied to disorders associated with anxiety, and ‘problem gambling’ has often been used to refer to a subdiagnostic threshold condition and therefore may unintentionally minimise the severity of the disorder.

Similarly to substance use disorders, the diagnostic criteria for gambling were examined with respect to their fit and classification accuracy in epidemiological and clinical samples. In DSM-IV, five of ten criteria are needed for diagnosis. The least prevalent criterion is the one related to committing illegal acts to gamble, such as forgery, fraud, theft or embezzlement, and this criterion is rarely endorsed in the absence of other criteria. ^{Reference Strong and Kahler9} Further, statistical analyses revealed that this item adds little to classification accuracy and it is therefore being recommended for deletion in DSM-5, similarly to the parallel criterion for substance use disorders.

The threshold for diagnosis was also considered. Independent analyses revealed that reducing the threshold to four criteria improved classification accuracy (e.g. Stinchfield ^{Reference Stinchfield10} ). Hence, the recommendation is that a gambling disorder diagnosis will constitute meeting four of nine criteria in DSM-5, and this change is unlikely to impact base rates substantially. ^{Reference Denis, Fatséas and Auriacombe11}

Other non-substance use addictive disorders

Several other putative conditions were considered for inclusion in this section. A subcommittee convened to assess the literature related to internet/gaming addiction, compulsive shopping, and several other behaviour patterns. The Substance-Related Disorders Work Group concluded that sufficient data do not yet exist to classify any of these conditions as mental disorders in DSM-5. Although a large literature exists on gaming addiction, no consensus exists regarding how the condition is best classified. Further, limited data are available regarding the prevalence or clinical course of the putative disorder. As more research is conducted, empirical data will guide decisions about whether and how best to diagnose such conditions for future editions of DSM, and gaming addiction is likely to be listed in the appendix as a potential disorder requiring more research.

How these changes may impact diagnosis and treatment in the UK

Despite the wide ranging use of DSM throughout the world, medical students in the UK are trained using ICD-10, and doctors across the country classify most psychiatric disorders using this manual. One exception relates to gambling. Although few physicians or researchers in the UK diagnose or treat gambling, physicians in practice as well as in the National Problem Gambling Clinic have officially adopted the DSM criteria for this disorder. Using the DSM criteria has allowed for greater collaboration in international research projects.

The changes in diagnoses with DSM-5 are unlikely to impact the classification of substance use disorders clinically or with respect to research in the UK, given the reliance on ICD-10 for these purposes. However, the changes proposed for gambling disorder, especially with regard to the placement of gambling disorder in a similar section with other addictive disorders, are generally welcome. The inclusion of gambling disorder in this section is likely to lend greater credibility to the concept of excessive gambling as a psychiatric disorder. Clinically, this change may encourage more screening efforts for gambling in primary care and substance misuse treatment clinics, which in turn may lead to greater recognition of the need for brief interventions for disordered gambling in high-risk patient populations. ^{Reference Petry, Weinstock, Morasco and Ledgerwood8} Greater awareness and recognition of gambling disorder can assist both clinicians and researchers to communicate better with policy makers, commissioners of services and research funders of the need to invest more in this field and expand treatment services to those with gambling disorder. Further, if the ICD follows the DSM lead on classifying gambling disorder alongside substance use disorders, physicians in training and clinicians in practice may be more likely to place it at par with substance use disorders. This change is long overdue.

Funding

Preparation of this report is based in part on National Institutes of Health grants R01-DA021567, R01-MH60417, P30-DA023918, R01-DA027615, R01-DA022739, R01-DA13444, R01-DA018883, R01-DA016855, R01-DA14618, P50-DA09241, P60-AA03510, and R01-DA024667.