BACKGROUND: Major depressive disorder (MDD) is a debilitating condition with significant economic consequences. Conservative estimates indicate that between 10% and 20% of all individuals with MDD are treatment resistant. The objectives for this study were (1) to use current treatment strategies identified in the literature to evaluate the validity of studying treatment-resistant depression (TRD) using claims data and (2) to estimate cost differences between TRD-likely and TRD-unlikely patients identified by use of treatment patterns. METHOD: The data source consisted of medical, pharmaceutical, and disability claims from a Fortune 100 manufacturer for 1996 through 1998 (N = 125,242 continuously enrolled beneficiaries between the ages of 18 and 64 years). The sample included individuals with medical or disability claims for MDD (NMDD = 4186). A treatment pattern algorithm was applied to classify adult MDD patients into TRD-likely (NTRD = 487) and TRD-unlikely groups. Resource utilization and costs were compared among TRD-likely and TRD-unlikely patients and a random sample of average beneficiaries (i.e., 10% of all beneficiaries) for 1998. RESULTS: Consistent with the epidemiologic literature, the algorithm classified 12% of the MDD sample as TRD-likely. Mean annual costs were $10,954 for TRD-likely patients, $5025 for TRD-unlikely patients, and $3006 for average beneficiaries. TRD-likely patients used almost twice as many medical services as did TRD-unlikely patients and incurred significantly greater indirect costs (p < .0001). CONCLUSION: It is feasible to use an administrative dataset to develop a claim-based treatment algorithm to identify TRD-likely patients. Resource utilization by TRD-likely patients was substantial, not only for direct treatment of depression but also for treatment of comorbid medical conditions. Additionally, TRD imposed on employers substantial indirect costs resulting from high rates of depression-associated disability.
BACKGROUND: Major depressive disorder (MDD) is a debilitating condition with significant economic consequences. Conservative estimates indicate that between 10% and 20% of all individuals with MDD are treatment resistant. The objectives for this study were (1) to use current treatment strategies identified in the literature to evaluate the validity of studying treatment-resistant depression (TRD) using claims data and (2) to estimate cost differences between TRD-likely and TRD-unlikely patients identified by use of treatment patterns. METHOD: The data source consisted of medical, pharmaceutical, and disability claims from a Fortune 100 manufacturer for 1996 through 1998 (N = 125,242 continuously enrolled beneficiaries between the ages of 18 and 64 years). The sample included individuals with medical or disability claims for MDD (NMDD = 4186). A treatment pattern algorithm was applied to classify adult MDDpatients into TRD-likely (NTRD = 487) and TRD-unlikely groups. Resource utilization and costs were compared among TRD-likely and TRD-unlikely patients and a random sample of average beneficiaries (i.e., 10% of all beneficiaries) for 1998. RESULTS: Consistent with the epidemiologic literature, the algorithm classified 12% of the MDD sample as TRD-likely. Mean annual costs were $10,954 for TRD-likely patients, $5025 for TRD-unlikely patients, and $3006 for average beneficiaries. TRD-likely patients used almost twice as many medical services as did TRD-unlikely patients and incurred significantly greater indirect costs (p < .0001). CONCLUSION: It is feasible to use an administrative dataset to develop a claim-based treatment algorithm to identify TRD-likely patients. Resource utilization by TRD-likely patients was substantial, not only for direct treatment of depression but also for treatment of comorbid medical conditions. Additionally, TRD imposed on employers substantial indirect costs resulting from high rates of depression-associated disability.
Authors: Jeffrey F Scherrer; Joanne Salas; Mark D Sullivan; F David Schneider; Kathleen K Bucholz; Thomas Burroughs; Laurel Copeland; Brian Ahmedani; Patrick J Lustman Journal: Prev Med Date: 2016-08-03 Impact factor: 4.018
Authors: Paul Greenberg; Patricia K Corey-Lisle; Howard Birnbaum; Maryna Marynchenko; Ami Claxton Journal: Pharmacoeconomics Date: 2004 Impact factor: 4.981
Authors: Howard G Birnbaum; Rym Ben-Hamadi; Paul E Greenberg; Matthew Hsieh; Jackson Tang; Camille Reygrobellet Journal: Pharmacoeconomics Date: 2009 Impact factor: 4.981