Validation of a method for identifying nursing home admissions using administrative claims

Abstract

Background: Currently there is no standard algorithm to identify whether a subject is residing in a nursing home from administrative claims. Our objective was to develop and validate an algorithm that identifies nursing home admissions at the resident-month level using the MarketScan Medicare Supplemental and Coordination of Benefit (COB) database.

Algorithms can be used to identify fragility fracture cases in physician-claims databases

“…Population-based administrative databases have been successfully used for case-identification of medical conditions in the general population [15][16][17][18][19][20][21][22][23]. Using administrative databases is relatively inexpensive, minimizes usual biases associated with survey studies (recall bias, nonresponse, and subject attrition), and increases external validity compared with cohort studies [15,21].…”

“…Physician-billing claims databases (PCDs) are a valid source of information for outcome assessment [15][16][17][18][19][20][21][22][23]. Since fractures are typically acute and painful, they are easily recognized and promptly treated within the healthcare system [17] and consequently, should be identifiable from PCDs.…”

Identifying types of nursing facility stays using medicare claims data: an algorithm and validation

“…Although previous studies identified nursing facility stays, they did not distinguish between short and long stays (Iwashyna 2003;Zuckerman et al 2007). Further, these studies were cross-sectional, with limited sample size or used claims based upon beneficiaries with supplemental insurance only.…”

“…Two studies have addressed the issue of detecting nursing facility use and nursing facility admissions in Medicare claims data (Iwashyna 2003;Zuckerman et al 2007). These studies did not differentiate between short-and long-term nursing facility care.…”

The usefulness of administrative databases for identifying disease cohorts is increased with a multivariate model

“…Zuckerman et al [16] described an algorithm to combine different types of provider claims to predict whether patients lived in a nursing home. Lix et al [17] combined information from hospital, physician, and pharmacy administrative data sets to predict the probability that patients had osteoporosis.…”