COMPARISON OF USING NDC OR ICD CODES TO SELECT PAIN PATIENTS USING DATABASE ANALYSES
Author(s)
Hall J1, Grossman P2, Trukhnova Z1, Huang J2, 1Ingenix, Eden Prairie, MN, USA; 2Purdue Pharma, Stamford, CT, USA
OBJECTIVES: Traditional database methodology suggest patient selection be done through ICD codes when interested in a specific disease and through NDC codes when interested in the medication. However, this methodology may be different for a population of pain patients, as no ICD code specific for pain exists. The purpose of this study was to understand the differences in data pulled using NDC versus ICD-9 codes to identify a pain population. METHODS: United Healthcare pharmacy and medical claims from continuously enrolled individuals in 1998 were used for examination. In one method patients were selected if they were at least 18 years of age as of January 1, 1998, were continuously enrolled with a drug and medical benefit and had at least one medical claim with a pain-related diagnosis (cancer, fibromyalgia, low back pain, neuropathic pain, and osteoarthritis) in 1998. The other method differed by selecting patients who had at least one pharmacy claim with an NDC for pain medication (long acting opioid, short acting opioid, combination opioid and NSAID). RESULTS: Using the ICD code method, 233,390 patients were selected. Low back pain had the highest amount of people, constituting 40%, while neuropathic pain had the least, 5%. The average age of patients with pain-related diagnosis was 46.6. Using the pharmacy claims method, 274,830 were selected. The most commonly prescribed pain medication was combination opioids (36%), and the least was long acting opioids (0.7%). The mean age for this population was 43.2. The overlap in patient selection was 84.9% with the prescription drug population being a larger cohort. CONCLUSIONS: Using NDC codes to select users of pain medications in a database revealed more users than those selected through ICD codes specific for conditions where pain exists. However, an 85% overlap in patient selection suggests that the diseases selected capture a majority of pain-related conditions.
Conference/Value in Health Info
2003-05, ISPOR 2003, Arlington, VA, USA
Value in Health, Vol. 6, No. 3 (May/June 2003)
Code
PNP5
Topic
Real World Data & Information Systems
Topic Subcategory
Health & Insurance Records Systems
Disease
Systemic Disorders/Conditions