Objective: To determine whether patients' hand-held immunization cards provide accurate assessments of immunization status when compared with their corresponding medical records.
Setting: Urban hospital emergency department immunization program.
Design: Comparison of 2 criterion standards.
Patients: Children aged 4 months to 6 years who presented consecutively with their immunization cards and received routine care in the hospital's pediatric clinic.
Selection: Of 673 eligible patients seen in the immunization program from November 1992 to October 1993, 140 were randomly selected for comparison of immunization card and medical record immunization dates; in addition, all 123 eligible patients seen between August and October 1994 were selected. Of the total of 263 children, medical records for 257 (98%) were available for review. The dates of diphtheria-tetanus-pertussis, polio, measles-mumps-rubella, and Haemophilus influenzae type b immunization from immunization cards and medical records were recorded, as were patient age, sex, and ethnicity. Immunization card-medical record immunization date pairs were compared. Each immunization card and medical record was categorized as up to date, due for immunization, or delayed 2 months or more for any immunization at the time of the visit.
Results: In 218 (85%) of 257 cases, the immunization card and medical record immunization dates were identical (McNemar test, P = .63). The immunization card and medical record agreed that patients were due for immunization in 91 cases and agreed that patients were not due for immunization in 138 cases (kappa = 0.77; 95% confidence interval, 0.70-0.85). The immunization card and medical record agreed that patients were delayed for 1 or more immunizations in 51 cases and agreed that patients were not delayed in 187 cases (kappa = 0.79; 95% confidence interval, 0.71-0.88).
Conclusion: The hand-held immunization card is a suitable alternative to the medical record when the need for immunization is assessed or when rates of immunization delay in populations are determined.