Comorbidities and Associated Treatment Charges in Patients with Anxiety Disorders

Trent McLaughlin, Ph.D., Erika C. Geissler, R.N., M.B.A., George J. Wan, Ph.D., M.P.H.


Pharmacotherapy. 2003;23(10) 

In This Article


In this analysis, the most common non-psychiatric comorbidity was hypertension, which was present in approximately 22% of study patients. This is a higher estimate than that of previous reports, which estimated the rate of comorbid hypertension in patients with mood disorders at 17%.[2] This difference may be related to the definition of mood disorders in the earlier analysis, as it involved conditions other than anxiety, as well as the time period; our analysis covered 1998-2000, versus 1996 in the earlier study.[2] The rates of comorbid diabetes and asthma were similar to those of previous research, although heart disease was more common in our study.[2]

The most common psychiatric comorbidity was depression, which was present in approxi-mately 37% of study patients, which is consistent with previously published estimates. For example, in the National Comorbidity Survey, patients with a diagnosed anxiety disorder had high rates of comorbid depression; 22.4% of patients had social phobia, 9.4% had agoraphobia, and 2.3% had panic disorder.[12]

Of note, the common psychiatric and nonpsychiatric comorbidities identified were ranked in the same order regardless of time period (before or after the initial anxiety diagnosis and anxiolytic prescription). Whether this trend should have been expected by design is uncertain.

In our study, the most-prescribed anxiolytic agents were benzodiazepines, such as alprazolam and lorazepam; concomitant antidepressant drug therapy was common. Our classification of anxiolytic agents was based on published sources.[13,14] Several newer antidepressant agents, such as sertraline, paroxetine, and extended-release venlafaxine, are now indicated for treatment of some anxiety disorders (e.g., posttraumatic stress disorder, panic disorder, social anxiety disorder, and generalized anxiety disorder). Further research is needed to determine the impact of treatment with these newer agents on trends in comorbidities and associated treatment charges.

Patterns of the treatment charges before and after initial anxiety diagnosis and anxiolytic prescription may be an artifact of our study design. Patients averaged approximately $9300 in treatment charges for the 12 months after initial anxiety diagnosis and anxiolytic prescription. This is approximately twice the estimate reported in an earlier study.[2] Again, the earlier estimate was based on patients with mood disorders, which included but was not restricted to anxiety.[2] Other factors that may explain the observed differences are the study periods (1996 vs 1998-2000) and the method of calculating treatment charges. The earlier study used the Medical Expenditure Panel Survey data, in which patients report their expenditures for health care. Our study used actual charge data from health care practitioners, which were submitted to receive payment for providing care.

The use of charges in our analysis may have inflated the cost estimate; charged amounts are typically higher than the amount reimbursed by insurers (which may be argued to be more representative of the true cost, since the nonreimbursed amount for any claim is typically not passed on to the patient or other payer). In addition, the estimate from the earlier study involved uninsured patients, who historically incur less cost than those covered by insurance.[2]

Our study, like any retrospective analysis of administrative claims data, has a number of limitations. First, identification of patients with anxiety disorder was based on medical claims with a diagnosis of anxiety. Because we could not confirm this diagnosis by reviewing patient charts or collecting any clinical test results, a number of patients may have been falsely identified as having anxiety when indeed they did not. To minimize the risk of such an occurrence, we required patients to have received anxiolytic pharmacotherapy. However, some patients still may have been misidentified.

Also of note, the measure used in our analysis was total charges; that is, all medical charges regardless of reason. Due to the complexity of anxiety symptoms and treatment, we did not attempt to measure anxiety-specific charges because this would lead to underestimation of the true economic impact of the disorder. Although this decision reduced the risk of missing anxiety-related costs, it conversely also meant that the differential between preperiod and postperiod charges we estimated may not be entirely due to anxiety, since other events may have led to the higher charges in the postperiod.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: