How many different ways do patients meet the diagnostic criteria for major depressive disorder?
Introduction
The era of specified inclusion and exclusion criteria to make psychiatric diagnoses began in 1972 when a group of researchers at Washington University published an article entitled “Diagnostic Criteria for Use in Psychiatric Research” [1]. Referred to as the Feighner criteria (after the lead author of the article), or the Washington University criteria (after the academic affiliation of the authors), this article delineated for the first time specific inclusion and exclusion criteria for 15 disorders that the authors considered to have been empirically validated. While this article is generally credited with ushering in the modern era of specified diagnostic criteria, in fact, this publication was not the first description of specific criteria for depression.
In a 1957 report on the clinical features of manic–depressive disorder, Cassidy and colleagues [2] diagnosed depressive disorder in patients who reported both low mood and 6 of a list of 10 symptoms. The ten criterion symptoms of depression listed by Cassidy et al. were slow thinking, poor appetite, constipation, insomnia, fatigue, loss of concentration, suicidal ideas, weight loss, decreased libido, and agitation. The Washington University definition of depression represented a modification of the Cassidy et al. symptom list, deleting one item (constipation) and adding others (guilt, worthlessness, indecisiveness, hypersomnia, pervasive loss of interest). Feighner et al. did not indicate the reasons for changing the criteria developed by Cassidy et al., and, to our knowledge, never published any data to empirically support their selection of particular symptoms over others. Nonetheless, the symptom inclusion criteria identified by the Washington University group have changed relatively little during the past 40 years, thus attesting to the astute observations of these clinical researchers. The symptom inclusion criteria for major depressive disorder (MDD) in the DSM-5 [3], as well as the threshold to distinguish between cases and noncases, are similar to the ones originally articulated by Feighner and colleagues.
The DSM-5 symptom inclusion criteria for MDD require the presence of at least five characteristic features from a list of nine, at least one of which must be low mood or anhedonia. This approach towards defining depression, based on a minimum number of features from a longer list, results in heterogeneity in patients’ clinical profile because there are many different possible combinations of criteria that qualify for a diagnosis. Altogether, there are 227 different ways to meet the symptom criteria for MDD, and it is possible for 2 patients diagnosed with MDD to have no symptoms in common. (Because some of the MDD criteria have multiple components, 2 patients can meet the same criterion in different ways. Thus, 2 patients can have no symptoms in common.)
While there are numerous ways in which a patient might meet the symptom criteria for MDD, it is also recognized that the criteria “hang together” as a syndrome thereby justifying the identification of the clinical syndrome. While the sensitivity and specificity of the MDD criteria vary, all are independently associated with the diagnostic construct [4], and they co-occur significantly beyond chance [5]. This raises the questions of whether all of the theoretically possible different ways of meeting the MDD criteria actually occurs in patients, and whether some combinations of criteria are much more common than others and capture the majority of patients diagnosed with MDD. In the present report from the Rhode Island Methods to Improve Diagnostic Assessment and Services (MIDAS) project, we examined in a large cohort of depressed patients the different ways in which they met the DSM-5 symptom inclusion criteria. We hypothesized that some of the 227 possible criteria combinations would occur more frequently than other combinations.
Section snippets
Methods
The Rhode Island MIDAS project represents an integration of research methodology into a community-based outpatient practice affiliated with an academic medical center [6]. A comprehensive diagnostic evaluation is conducted upon presentation for treatment. This private practice group predominantly treats individuals with medical insurance (including Medicare but not Medicaid) on a fee-for-service basis, and it is distinct from the hospital’s outpatient residency training clinic that
Results
Of the 1566 patients who met criteria for MDD, the mean number of criteria met was 6.8 (S.D. = 1.3), and the modal number of criteria met was 7 (25.2%, n = 394). One hundred fifty-seven individuals met all 9 criteria (10.0%), and 19.6% (n = 307) met the minimum number of 5 criteria.
Depressed mood was the most frequent criterion (93.7%) and suicidal thoughts the least frequent (50.0%) (Table 1). The rank order of criterion frequency was similar for patients who met 5, 6, 7 or 8 MDD criteria (Table 1
Discussion
Polythetic definitions of psychiatric disorders, which require a minimum number of criteria from a list, potentially result in diagnostic heterogeneity because there are many different ways to meet criteria. However, the diagnostic criteria used to define a disorder are not randomly occurring signs and symptoms but instead “hang together” and co-occur at greater than chance levels, thus concerns about diagnostic heterogeneity may be more theoretical than actual. To the best of our knowledge
References (11)
- et al.
Diagnostic criteria for use in psychiatric research
Arch Gen Psychiatry
(1972) - et al.
Clinical observations in manic-depressive disease. A quantitative study of one hundred manic–depressive patients and fifty medically sick controls
JAMA
(1957) Diagnostic and Statistical Manual of Mental Disorders.
(1994)- et al.
Diagnosing major depressive disorder I: a psychometric evaluation of the DSM-IV symptom criteria
J Nerv Ment Dis
(2006) - et al.
Diagnosing major depressive disorder: II: is there justification for compound symptom criteria?
J Nerv Ment Dis
(2006)
Cited by (0)
Potential conflicts of interest: None.