American Journal of Public Health · 1993 · 32 citations · 4 references
Health ProfessionsBiomedical InformaticsHealth InformaticsClinical Classification SystemsElectronic Health RecordsMedical CodingData CodingElectronic Health Record CodingHealthcare Coding SystemsHiv DiagnosisHiv SurveillanceMedical Coding SystemsAcquired Immunodeficiency SyndromeClinical Coding AccuracyDiagnosis CodingHealth Record CodingHuman Immunodeficiency Virus SurveillanceIcd CodingMedical Record AccuracyPositive Predictive Value
To evaluate the accuracy of computerized medical‑record coding for human immunodeficiency virus (HIV). Medical charts were reviewed in six sites. The study found that HIV codes had a predictive value of 91% or higher, identified HIV in 34% of Medicaid records lacking an HIV code, and that AIDS was detected in both HIV‑coded and AIDS‑coded records, though undercoding of HIV occurred in Medicaid data.
To evaluate the accuracy of computerized medical-record coding for human immunodeficiency virus (HIV), medical charts were reviewed in six sites. In 7601 hospital and 867 Medicaid records with a listed diagnosis of HIV, the predictive value for HIV was 91% or higher. HIV was identified in 34% of 1155 Medicaid records listing immune disorder or illness in the acquired immunodeficiency syndrome (AIDS) surveillance definition (without an HIV code). In hospital and Medicaid records, AIDS was identified both in records listing AIDS and records listing HIV without AIDS. HIV codes on hospital and Medicaid records were highly predictive for HIV; undercoding of HIV occurred in Medicaid records.
4
The 'Medicaidization' of AIDS. Trends in the financing of HIV-related medical care
The 1987 US Hospital AIDS Survey
Dennis P. Andrulis · JAMA · 1989 · 62 citations