WISQARS™ Cost of Injury Reports Help Menu
5.2 Source and Coding of National Data for Injury-Related Hospitalizations and ED Visits
Nationally representative data describing injuries that resulted in hospitalization subsequent to emergency department (ED) treatment or that resulted in an ED visit followed by release (without hospitalization or other transfer) were obtained from the National Electronic Injury Surveillance System – All Injury Program (NEISS-AIP) for calendar year 2010. NEISS-AIP is an extension of the National Electronic Injury Surveillance System (NEISS) operated by the U.S. Consumer Product Safety Commission (CPSC). NEISS-AIP is a collaborative effort between CPSC and the National Center for Injury Prevention and Control (NCIPC).
Since May 1971, CPSC has used NEISS, or a precursor of it, to monitor consumer product-related injuries as part of its regulatory jurisdiction. As of the year 2000, NEISS collected information from a nationally representative sample of 100 U.S. hospital EDs. The participating hospitals represent a stratified probability sample of all U.S. hospitals (including those in U.S. territories) that have at least six beds and provide 24-hour emergency services. NEISS-AIP began operating in July 2000 at a nationally representative subset comprised of 66 of these 100 hospitals. NEISS and NEISS-AIP hospitals include very large inner-city hospitals with trauma centers as well as large urban, suburban, rural, and children’s hospitals.2
Trained coders working at NEISS-AIP hospitals abstract data on all injuries or poisonings treated in the ED. The data are abstracted from ED records and include the intent of injury, primary body region affected, principal diagnosis (i.e., nature of injury), sex and age of patient, and other selected characteristics. The coded data and a narrative description for each case are then entered into an on-site computer and electronically transmitted to CPSC headquarters in Bethesda, MD. NEISS-AIP quality assurance staff at CPSC headquarters review the coded data elements and narrative for each case, and use this information to assign codes indicating the precipitating and direct causes of injury and also to confirm the coder-assigned injury intent for each case. WISQARS Cost of Injury Reports refers to the precipitating cause to classify the mechanism (cause) each of injury case.
In most instances, a case narrative provides enough detail about the circumstances of the injury event to allow the quality assurance staff at CPSC headquarters to assign codes for the precipitating and direct causes of injury and to confirm the intent of injury assigned by the hospital coder. The on-site hospital coders are trained not to interpret information provided in the ED records when writing the narratives but rather to transcribe exactly what the medical providers recorded about the ED visits. Most narrative descriptions are based on doctors’ and nurses’ notes. This practice helps ensure that the data from the narratives are accurate and consistent. In a joint effort, CPSC and NCIPC developed coding guidelines for intent and mechanism (cause) of injury. These guidelines are consistent with coding guidelines in the International Classification of Disease (ICD) 9-CM.8,15 For further information about coding procedures, see the NEISS Coding Manual.16 NEISS-AIP currently provides data on over 500,000 injury-related ED cases yearly. Data for these cases are weighted by the inverse of the probability of (hospital) selection to support the calculation of national estimates.
Note: Stratification of data using the intent by mechanism classification scheme or the body region by nature of injury classification scheme, and/or stratification of data by sex and age may result in reports with some strata in which the case counts are not large enough to support statistically reliable cost estimates. National cost-of-injury estimates that are based on 20 or fewer injury surveillance records or based on a national (weighted) estimate of 1200 or fewer cases are considered statistically unreliable (similar to the reliability/stability criterion employed in other uses of NEISS-AIP data). Even with sufficiently large case counts (surveillance / estimated) some cost-of-injury estimates can be statistically unreliable due to high variability of the case-level costs. For additional details on how cost estimates are evaluated in terms of statistical reliability, refer to Section 4.5.