Impact of Pneumococcal Conjugate Vaccines on Otitis Media and Acute Otitis Media in Swedish Children

A step-wise approach will be applied: 1. Descriptive analyses to assess any trends in OM and AOM incidence, any seasonality (repeating pattern of incidence related to time such as seasons, quarters and months), the order of seasonality, extreme observations/outliers, whether incidence trends include abrupt changes. 2. If trends or seasonality are identified, a time series analysis will be conducted in 2 steps. a. Time series analysis of monthly incidence, using smoothing technique to estimate differences in pre-PCV, PCV7, and Synflorix®/PCV13 vaccination periods. b. Interrupted time-series analysis if the above steps show any differences in pre- and post- vaccination introduction periods. Diagnoses of AOM recorded under the ICD-10 code H66 (suppurative and unspecified OM) and ICD-10 H66.0 (acute suppurative OM); Chronic OM under ICD-10 code H66.3 (other chronic suppurative OM) and other OM underICD-10 code H66.9 (OM unspecified) and H65 (non-suppurative OM).

The impact of PCV will be investigated by looking at trends over time in the rate of antibiotics* dispensed due to an OM or AOM diagnosis for the two counties separately. The analysis of rate of antibiotic use will be performed in two steps as: Descriptive statistics for the number of dispensed antibiotics. Time series analysis for the incidence of dispensed antibiotics. *Date of an antibiotic dispensed within 7 days of a registered diagnosis code (ICD-10 code H65, H66, H66.0, H66.3, H66.9).

The rate of tympanostomy tube insertions* among children with OM or AOM will also be analyzed in the same way as the analysis of frequency of antibiotics dispensed. *Date of Tympanostomy tube placement procedure (KVA DCD10).

The impact of PCVs on the incidence of severe AOM among children will be analysed with a step-wise time series analysis approach, which is described for the primary outcome. Severe AOM defined as: -Children diagnosed with AOM with presence of moderate to severe otalgia or fever ≥ 39° C, or -Children with ≥3 diagnoses of AOM in a 6-month period or ≥4 AOM diagnoses in a 12-month period, or-Children reported with hospitalization due to AOM or any documented evidence of IV infusion of antibiotics in hospital.

Cost dispersion statistics (mean, median, standard deviation, minimum, maximum, variance) will be calculated. Direct costs for OM and AOM will be calculated by including healthcare resource utilization for each child as the count of antibiotic prescriptions filled for OM and AOM, and use of medical services, such as visits to the general practitioner, visits to the hospital, procedure costs (tympanostomy tube placements, myringotomies and IV antibiotics administrations) and length of hospital stay during the follow-up period. Cost for all health care visits and procedures will be calculated using the number of registered events multiplied by a unit cost based on public price lists for that particular visit or procedure.

The following analysis will be conducted to understand individual-level factors that impacts OM and AOM rates and time to first incidence of an OM or an AOM separately for each county council. A survival analysis will be conducted to assess predictors of the time to first OM or AOM using a left censored Cox proportional hazards model.