Evaluation of causes-of-death: which statistics should we rely on, hospital deaths or vital statistics?
Sutra, S.; Chirawatkul, A.; Leelapanmetha, P.; Sirisuwan, S.; Thepsuthammarat, K.
Journal of the Medical Association of Thailand 95 Suppl. 7: S262-S273
2012
ISSN/ISBN: 0125-2208 PMID: 23130464 Document Number: 500296
Age-specific causes of death yield important information for planning health services and medical education. To compare the age-specific causes of death between in-hospital deaths and death registration statistics. Information on in-hospital mortality in fiscal 2010 was extracted from the three health insurance schemes. Death registrations (ICD-10 coding) were from the Bureau of Registration Administration, Ministry of Interior Statistics on the age-specific causes of death were analyzed. In-hospital deaths numbered 132,512 (47.5% occurring in tertiary care) vs. 411,331 recorded in vital statistics (68% died outside hospitals). Most (74%) infants died in-hospital so causes were clearly documented vs. death registration. A minority (6.2%) of in-hospital deaths and of death registrations (9.7%) were due to unnatural causes. The majority (79.5%) of unnatural deaths died before arriving at hospital. Ill-defined codes for causes of death were found in 6.1% of in-hospital records and 42.2% of death registrations. After censoring ill-defined codes, the ten leading, age-specific causes of death agreed between the two data sets. Medical personnel should receive training to do proper death certification. Periodic validation of hospital COD certification should also be done. Use of verbal autopsy in the Thai context would help to standardize record-keeping and to reduce ill-defined codes for deaths occurred outside hospital.