The methodology of the SDIS is a modification of the methods used by the NIS.6
Data obtained from the SDIS are collected in two phases: a telephone survey followed by verification of immunization information when necessary.
In the first phase, a random-digit-dialing (RDD) telephone survey is conducted annually between March and June. The RDD methodology is used to identify households in San Diego that have a 19- to 35-month-old child. When a household with an appropriately aged child is identified, demographic information about the household is obtained from the RDD respondent. Eighty-five percent (711/839) of the 2005 SDIS respondents were the mothers of the age-eligible children identified in the household. If the respondent has a state-issued written immunization record for the child, he/she is asked to read from that record the dates when the child received doses of diphtheria, tetanus toxoids, and acellular pertussis (DTaP), polio, measles-mumps-rubella (MMR), Haemophilus influenzae type b (Hib), hepatitis B (HepB), varicella (VAR), pneumococcal conjucate vaccine (PCV), hepatitis A (HepA), and influenza (flu) vaccine. The respondent is also asked to read the name and contact information for the provider that administered each dose. If the written immunization record used during the RDD interview indicates that the child was fully vaccinated, the SDIB considers the report to be a sufficient documentation of immunization history and further information is not requested from either the child's vaccination provider or the registry.
If an immunization record is not available, then the number of doses is recorded from the respondent's memory. In the 2005 SDIS, a child was considered to be fully vaccinated if information from the immunization record showed that the child had received doses approximately at the recommended ages and intervals of ≥4 doses of DTaP, ≥3 doses of polio, ≥1 dose of MMR, ≥3 doses of Hib, ≥3 doses of HepB, ≥1 dose of VAR (or history of varicella disease), ≥3 doses of PCV, 1 dose of HepA if the child was ≥24 months of age or 2 doses of HepA if the child was ≥30 months of age and more than six months had passed since his/her first dose, and ≥1 dose of flu vaccine. If the immunization record did not indicate that one dose of flu vaccine was administered, then the respondent's recall was accepted. Surveys in years prior to 2004 do not include all of these vaccines. As new vaccines become available and recommended, the SDIB adds them to the list of vaccines that it monitors through the SDIS.
The second phase of data collection is conducted if the RDD respondent's report of the child's vaccination history is not read from an immunization record or if data from the immunization record do not show that the child is fully vaccinated. Consent to contact the child's provider is requested verbally from all respondents during the telephone interview. If consent is obtained and second-phase data collection is warranted, the San Diego Regional Immunization Registry (SDIR) is queried to verify the child's vaccination history. If a complete immunization record cannot be acquired using SDIR, then it is requested from the child's medical provider(s). The SDIB attempts to contact local, out-of-state, and out-of-country providers for verification. Although providers typically comply with the request for verification, some ask for documentation. In these situations, SDIB sends a form with the interviewer's signature on it stating that verbal consent was received from the respondent during the telephone interview. If the provider insists on receiving written consent from the respondent, SDIB is not able to obtain the information necessary to verify immunization status.
Survey data collected by the SDIS are weighted to represent the target population of children 19 to 35 months of age living in San Diego County. The statistical weighting methods account for the probability of being sampled, households that do not have a telephone, and nonresponse from medical providers. Initial weights are constructed using poststratification so that the distribution of the weighted sample with respect to the selected demographic characteristics is identical to the same distribution in the population in San Diego. Nonresponse from medical providers is handled in the same way it is handled in the NIS.7