Consensus development and application of ICD-9-CM codes for defining chronic illnesses and their complications

Background: One particularly difficult challenge in evaluating disease management (DM) programs is defining the scope of economic outcomes to include in the evaluation. Measuring 'all-cause utilization' or 'total costs' assumes that a DM intervention impacts the entire spectrum of services rendered and reduces total medical costs, while limiting the evaluation to 'disease-specific' costs of the conditions under management may fail to capture any effect the program may have on complications directly related to that primary condition. An acceptable compromise between the two options is to include costs associated with diagnostic codes for the primary condition and those of medical complications directly related to that condition. Objective: To develop consensus on the International Classification of Diseases, ninth revision, Clinical Modification (ICD-9-CM) codes defining the primary conditions and complications of coronary artery disease (CAD), congestive heart failure (CHF), asthma, and chronic obstructive pulmonary disease. Methods: A modified Delphi technique, involving two panels of three physicians each (one consisting of cardiologists and the other of pulmonologists) and a physician consultant, was conducted via email and used to establish 100% consensus on the ICD-9-CM codes to be included in order to capture the appropriate costs for each of the primary conditions considered and their complications. The codes for primary conditions included by the panel were compared with those included in industry references. Results: Total consensus on the codes to be included for each of the primary conditions was reached within three rounds. Near-consensus on the codes to be used for complications for conditions was reached after the first round; however, four additional rounds were required for total consensus. Regarding the primary conditions, greatest agreement between the codes included by the panel and the various industry references was seen for asthma, with poor agreement observed between sources of codes for CAD and CHF. Conclusion: It is suggested that these lists of ICD-9-CM codes developed by consensus be used in evaluations across the industry to define the utilization and/or costs associated with DM interventions. The consistent use of these codes will greatly strengthen the validity of the current evaluation approach and consequently substantiate the value proposition offered by the industry.