BACKGROUND: Researchers conducting cost-outcome studies must account for all materially relevant care that subjects receive from their care providers. However, access to provider records is often limited. This article describes and tests the Utilization and Cost Inventory (UAC-I), a structured patient interview designed to measure costs of care when access to provider records is limited. METHODS: UAC-I was tested on 212 consenting adult veterans with mood disorder attending a VA medical center. Counts (inpatient days and outpatient encounters) and costs (dollars) computed from survey responses were compared with estimates from medical records and an alternative structured questionnaire. RESULTS: The agreement between inpatient costs computed from provider records and from UAC-I responses, assessed using the intraclass correlation coefficient (ICC), was 0.66, 95% confidence interval (CI), 0.30-0.84; the bias was -3.7%, 95% CI, -48 to 41. The ICC for the service data (inpatient days) was 0.97, 95% CI, 0.95-0.99; the bias was <1%, 95% CI, -14 to 15. The ICC for outpatient costs computed from provider records and from UAC-I responses was 0.53 95% CI, 0.38-0.65; the bias was <1%, 95% CI, -27 to 27. The ICC for outpatient encounters was 0.74, 95% CI, 0.65-0.80; the bias was <1%, 95% CI, -16 to 18. CONCLUSIONS: These results indicate that it may be feasible for cost-outcome studies to compare patient groups for inpatient and outpatient costs computed from patient self-reports.
BACKGROUND: Researchers conducting cost-outcome studies must account for all materially relevant care that subjects receive from their care providers. However, access to provider records is often limited. This article describes and tests the Utilization and Cost Inventory (UAC-I), a structured patient interview designed to measure costs of care when access to provider records is limited. METHODS: UAC-I was tested on 212 consenting adult veterans with mood disorder attending a VA medical center. Counts (inpatient days and outpatient encounters) and costs (dollars) computed from survey responses were compared with estimates from medical records and an alternative structured questionnaire. RESULTS: The agreement between inpatient costs computed from provider records and from UAC-I responses, assessed using the intraclass correlation coefficient (ICC), was 0.66, 95% confidence interval (CI), 0.30-0.84; the bias was -3.7%, 95% CI, -48 to 41. The ICC for the service data (inpatient days) was 0.97, 95% CI, 0.95-0.99; the bias was <1%, 95% CI, -14 to 15. The ICC for outpatient costs computed from provider records and from UAC-I responses was 0.53 95% CI, 0.38-0.65; the bias was <1%, 95% CI, -27 to 27. The ICC for outpatient encounters was 0.74, 95% CI, 0.65-0.80; the bias was <1%, 95% CI, -16 to 18. CONCLUSIONS: These results indicate that it may be feasible for cost-outcome studies to compare patient groups for inpatient and outpatient costs computed from patient self-reports.
Authors: Adam H Miller; Paul E Pepe; Ron Peshock; Rafia Bhore; Clyde C Yancy; Lei Xuan; Margarita M Miller; Gisselle R Huet; Clayton Trimmer; Rene Davis; Rebecca Chason; Micheal T Kashner Journal: Acad Emerg Med Date: 2011-05 Impact factor: 3.451
Authors: T Michael Kashner; Madhukar H Trivedi; Annie Wicker; Maurizio Fava; Stephen R Wisniewski; A John Rush Journal: CNS Neurosci Ther Date: 2009-08-27 Impact factor: 5.243
Authors: T Michael Kashner; Madhukar H Trivedi; Annie Wicker; Maurizio Fava; Kathy Shores-Wilson; Stephen R Wisniewski; A John Rush Journal: Int J Methods Psychiatr Res Date: 2009-09 Impact factor: 4.035
Authors: Vicky Stergiopoulos; Agnes Gozdzik; Jason Tan de Bibiana; Tim Guimond; Stephen W Hwang; Donald A Wasylenki; Molyn Leszcz Journal: BMC Health Serv Res Date: 2016-08-24 Impact factor: 2.655