BACKGROUND: Newly diagnosed cases of notifiable conditions are required to be reported to the state, including case vital status. The three notifiable conditions most commonly reported in association with death are Influenza-Associated Death, Streptococcus Group B Invasive Disease (GBS), and Streptococcus pneumoniae(Invasive). Despite the reporting requirements, it is unknown how many fatal cases of these diseases are not reported. Death certificates are a source of data that may be useful to identify unreported fatalities from these diseases. The objectives of this study were to identify whether cases of specific reportable diseases with reported deaths were correctly reported on death certificates; and to identify if reportable diseases were listed on death certificates as causes of death, but not reported.
METHODS: Data collected from Indiana’s National Electronic Disease Surveillance System (I-NEDSS) and the Indiana Death Registration System (IDRS) for deaths occurring from 2012 to 2014 was used to search for notifiable conditions. Capture-recapture analyses were used to estimate the number of expected confirmed fatal cases of these diseases in Indiana, and estimate the proportion of cases that were actually reported in I-NEDSS. Information from death certificates was identified by searching International Classification of Diseases Version 10 (ICD-10) mortality codes specific to the study diseases.
RESULTS: A total of 374 death certificates and 494 case investigations were identified to find 147 matches. Of these, 128 (87%) were influenza and 19 (13%) were Streptococcus pneumoniae. No matches were identified for GBS. Capture-recapture analysis estimated 1,257 fatal cases of influenza, Streptococcus pneumoniae, and GBS would be expected during the three-year study period (95% CI: 1,125 – 1,390). I-NEDSS identified 39% of the total estimated cases, while IDRS only identified 30% of the total estimated cases. The inclusion of IDRS-identified cases increased identification of the total estimated cases to 57%. For influenza fatalities, I-NEDSS had 43% sensitivity, while IDRS had 66% sensitivity. For Streptococcus pneumoniaefatalities, I-NEDSS had 27.5% sensitivity, while IDRS had 8% sensitivity. GBS has no matched fatalities and sensitivity could not be calculated.
CONCLUSIONS: Fatalities associated with notifiable diseases were found in both systems; however they did not always match up. Streptococcus pneumoniae fatalities were three times more likely to be reported in I-NEDSS; while influenza fatalities were more likely to be reported on death certificates. Use of death certificate information would provide a more complete analysis of fatal disease burden due to influenza, Streptococcus Group B, and Streptococcus pneumoniae in Indiana.