Capture-recapture and anti-tuberculosis drug prescriptions, Balearic Islands, Spain, 2010-2012.

SETTING: Tuberculosis (TB) control requires accurate knowledge of TB incidence, but few studies have been published on TB incidence using individual data from drug prescriptions. OBJECTIVE: To measure the estimated completeness of regional surveillance system data before and after adding data from anti-tuberculosis drug prescriptions. DESIGN: We compiled data on all individuals who received specific anti-tuberculosis drugs from three sources-the passive surveillance system, the active surveillance system and the anti-tuberculosis drug prescription database-in the Balearic Islands, Spain, between 2010 and 2012. We applied the capture-recapture method to measure completeness of data reporting. Statistical significance was set at P < 0.05, and 95% confidence intervals (CIs) were calculated. RESULTS: We detected 605 incident cases, compared with an estimated incidence of 719 cases (95%CI 646-793), representing 84.1% data completeness. We detected 480 of an estimated 559 pulmonary TB cases (95%CI 501-617; 85.9% data completeness) and 125 of an estimated 161 extra-pulmonary TB cases (95%CI 113-211; 77.6% data completeness). CONCLUSIONS: By adding a new source, we increased detection with respect to a previous study by 10.4% for all forms of TB, 9.7% for pulmonary TB and 13% for extra-pulmonary TB.

Estimated incidence of tuberculosis by capture-recapture in the Balearic Islands, 2010-2012.

SETTING: The Balearic Islands, Spain, 2010-2012. OBJECTIVE: To assess the completeness of TB surveillance, and to examine observed and estimated incidence rates (IR) compared with the period 2005-2007. DESIGN: The completeness of TB surveillance was assessed using the capture-recapture method. Confidence intervals were calculated using log-linear models. Data sources were records of mandatory notifiable diseases, active hospital surveillance and computerised primary health care history. RESULTS: Surveillance completeness for respiratory TB (R-TB) was 99.5%; it was 100% for non-respiratory TB (NR-TB). The overall observed IR was 16.32 cases/100 000 for all types of TB, 3.4 for NR-TB and 12.9 for R-TB. The estimated IR of TB all forms was 16.35/100 000 (95%CI 16.26-16.53), for NR-TB it was 3.4/100 000 (95%CI not calculated) and for R-TB it was 13.0/100 000 (95%CI 12.85-13.12). Surveillance completeness for R-TB during the period 2005-2007 was 65.2%; the observed IR was 22.6/100 000 and the estimated IR 31.3/10 000 (95%CI 25.0-37.7). CONCLUSION: The similarity of observed and estimated IR during the period 2010-2012 indicates the validity of the surveillance and the use of improved systems such as electronic medical records in recent years.