THE ECONOMIC BURDEN OF INVASIVE PNEUMOCOCCAL DISEASE IN ONTARIO, CANADA: A POPULATION-BASED STUDY
To determine incidence-based healthcare costs attributable to invasive pneumococcal disease (IPD) from the healthcare payer perspective.
We conducted a population-based cohort study of residents in Ontario, Canada (population 13 million) using individually linked laboratory and health administrative data. Using Public Health Ontario's laboratory information system database, we identified incident subjects with laboratory-confirmed IPD between 01/01/2009 and 31/12/2013 and followed them until 31/03/2014. Infected subjects were matched without replacement to uninfected subjects in a 1:3 ratio using propensity score and hard matching. Attributable costs were calculated by subtracting the mean costs of the matched uninfected subjects from those of each infected subject. The phase-of-care approach was employed to estimate costs over the disease course, organizing each subject’s observation time into 3 phases: acute infection, continuing care and end-of-life. We determined the length of acute infection and end-of-life care to be 30 days based on joinpoint analysis and expert opinion, while the remaining observation time was assigned to continuing care. Subsequently, survival-adjusted costs were calculated by applying the phase-of-care costs to crude survival probabilities.
We successfully linked 2,452 subjects with laboratory-confirmed IPD to the health administrative data. This population had a mean age of 57.1 (SD±24.2) and was 54.4% male. Of these patients, 89.5% were hospitalized during acute infection, while the all-cause mortality over the follow-up period was 33.1%. Mean follow-up was 870 days. Mean attributable 10-day costs were $5,574 (95% confidence interval [CI]: $5,351; $5,806) for the acute infection and $111 (95%CI: $105; $118) for continuing care. Attributable costs for the acute infection were lower for patients <18 years of age and significantly greater for those with high levels of comorbidity. Infected subjects who died in the first 30 days of infection had significantly higher end-of-life costs than their matched controls with mean attributable 10-day costs of $14,207 (95%CI: $8,965; $19,449). One-year survival adjusted-costs were $17,446 (95%CI: $15,031; $19,862) per patient.
IPD is associated with a substantial economic burden, leading to increased acute and long-term healthcare costs compared to general population controls. This represents the first study to evaluate the costs of IPD among patients with laboratory-confirmed disease in Canada, and will serve to inform the economic evaluation of future pneumococcal vaccination strategies.