Determining indications for adult vaccination: patient self-assessment, medical record, or both?

Context: Eight or more vaccines may be indicated for adults in the United States. Determining if any vaccines are needed requires integrating information on the patient's demographic and behavioral risk factors and health status, the health status of the patient's close contacts, and the patient's immunization history. This process can be time consuming for providers and their staff. We used patient self-assessment as a method of determining which vaccines are indicated for a patient and whether indicated vaccines had been received.

Design, Setting, And Participants: Cross-sectional convenience sample of 300 adults in three family practice settings. Participants completed a self-assessment tool to determine if influenza, pneumococcal, measles, mumps, and rubella (MMR), tetanus, hepatitis A and hepatitis B vaccines were indicated and previously received. A chart audit was then performed to obtain similar information.

Main Outcome Measures: Agreement (kappa statistic [< 0.00: poor agreement; 0.00-0.20: slight; 0.21-0.40: fair; 0.41-0.60: moderate; 0.61-0.80: substantial; 0.81-1.00: almost perfect]) between the self-assessment tool and the audit for (1) indicated vaccines and (2) previous receipt of indicated vaccines indicated according to both the assessment form and the audit.

Results: Agreement between the self-assessment tool and chart review was substantial or better only for pneumococcal and MMR vaccines (kappa=0.65 and 0.85, respectively). For influenza vaccine, agreement improved (from kappa = 0.56 to kappa = 0.74) when indications attributable to health conditions of family members were excluded. Agreement regarding receipt of vaccines was highest for influenza vaccine (kappa = 0.70). Only 57% of patients correctly recalled tetanus vaccination that were documented in the medical record (kappa = -0.04). Kappa statistics were unreliable for hepatitis A and B vaccines because so few vaccinations had been received.

Conclusions: Discrepancies in agreement regarding indications for vaccines appeared to result from absence of information in the medical record regarding high risk behaviors and family contacts. Lack of agreement regarding vaccines that had been previously been received appeared due to both poor recall and lack of documentation. Combining medical record audit with self-assessment may be the most complete assessment of vaccination status of adults, but requires reconciling disagreements. Electronic medical records and registries that contain information about risk factors and previously administered vaccines may be necessary to overcome some these problems.