![[ Visit Client Website ]](images/banner.png)
BACKGROUND: The New York City (NYC) Bureau of Tuberculosis Control (BTBC) tracks deaths among reportable TB events within its electronic case management and surveillance system (Maven), and obtains death certificates for all decedents as standard protocol. However, the contribution of TB to the cause of death is not thoroughly examined as recommended by the Centers for Disease Control and Prevention. We sought to evaluate the completeness and reliability of BTBC surveillance data associated with TB-related deaths in 2011.
METHODS: Death certificates from 2011 listing TB as a cause or contributor to death were obtained from the NYC Office of Vital Records. Data for reportable TB events, including cases, suspects, and contacts, with a death in 2011 were abstracted from Maven. Decedents from both data sources were matched using demographic characteristics to identify overlaps and discordant information. Data were then reanalyzed to determine if TB was a contributing cause of death. Any death where TB or TB therapy was documented as a cause of death in all available data sources was considered TB-related. Among those with a TB-related death, demographic and clinical characteristics were described by calculating means and proportions. The reliability of Maven data was examined by assessing the completeness of key death-related variables.
RESULTS: In 2011, 32 deaths were TB-related: 28 (88%) were listed in both data sources, and 4 (13%) were not reported to BTBC and were found only in vital records. An additional 28 deaths were identified as potentially TB-related, but a final determination of whether TB contributed to death could not be made without another source of death data. Among the 32 TB-related deaths, the mean age at death was 63 years, and a predominant proportion were male (78%), Asian (38%), and foreign-born (78%). Among the 28 TB-related deaths listed in Maven, 26 (93%) were confirmed as TB cases within one year of death; 7 (27%) of these had multiple sites of TB disease. There was variation in the completeness of death-related information among these 28 deaths; 100% had a date of death noted but only 71% had a cause of death recorded.
CONCLUSIONS: While this evaluation confirmed that BTBC captured most TB-related deaths in 2011, 13% were not identified in Maven; further evaluation of this is required. Completeness of death-related data in Maven also varied. These findings highlight areas for surveillance enhancement such as improving cause of death documentation and reviewing vital records to identify unreported TB cases.