Table 4 NEISS-AIP limitations identified by authors of NEISS-AIP articles, 2001–2021 (n = 167)
NEISS-AIP limitation | Number of articles |
|---|---|
Only provides national estimates therefore, state and local estimates are not available | 38 |
Doesn’t capture sufficient information on protective factors and injury circumstances | 35 |
Only captures information on most severe injury diagnosis or injured body part* | 32 |
Limited and variable amount of data details captured in narratives | 20 |
No injury outcome information | 18 |
Lack of information on injury severity | 13 |
Missing detailed location information (e.g., room in home where injury occurred) | 13 |
Small sample size can result in unstable estimates for some injuries* | 12 |
No measures of socioeconomic status (e.g., health insurance coverage) or other psycho-social variables (e.g., drug or alcohol use)* | 11 |
High level of missing race/ethnicity data | 7 |
Short (two-line) narrative captures limited information* | 5 |
Limited data available in the ED record (e.g., use of protective equipment) which is used as the source of NEISS-AIP data | 4 |
Body part injured not specific enough as only body region captured | 4 |
Data not linked to other data sources for verification (e.g., police reports) | 3 |
ICD-9-CM codes not captured | 3 |
Unknown and unintentional injury intents grouped together with potential for misclassification | 3 |
Injury intent difficult to identify in emergency department data (e.g., for interpersonal violence and self-inflected injuries) | 3 |
No data on sexual orientation or gender identity** | 2 |
No product codes available for more recently introduced consumer products | 1 |