Print document
TITLE VALIDATION STUDY OF VERBAL AUTOPSY OF CAUSES OF DEATH
AUTHOR TUBTIM TONGWIJIT
DEGREE MASTER OF ARTS PROGRAM IN POPULATION AND SOCIAL RESEARCH
FACULTY INSTITUTE FOR POPULATION AND SOCIAL RESEARCH
ADVISOR PRAMOTE PRASARTKUL
CO-ADVISOR SUREEPORN PUNPUING
 
ABSTRACT
Due to defects in the reliability of Thailand’s causes of death data, formulation of the public health policy and planning for curative and preventive strategies is not fully responsive to people’s actual health needs. In this study , there are three methods of establishing the cause of death : (1) diagnosed by medical specialist, the gold standard ; (2) verbal autopsy (VA), a technique where by public health personnels establish the cause of death through interviews with the deceased’s relatives and friends;(3) report of the cause of death from death certificate, almost by local registrars.The objective of this thesis is to assess the sensitivity and specificity of the VA technique against the gold standard and to compare its consistency against the third method. In this study, 4,050 deaths in Nakorn Sri Thammarat during July 1, 1997–June 30, 1998, were studied by using the VA. It was found that about 2,339 cases or 58 percent were identified causes of death by VA technique. Data without diagnosis causes of death by public health personnels accounts for 1,711 cases or 42 percent. It was found that the majority were diseases of circulatory system (such as cerebrovascular diseases, hypertensive diseases and ischemic heart diseases), ill-defined causes (including senility), and external causes of morbidity and mortality (such as transport accidents and assault). It was found that the VA technique best fitted ill-defined causes, including senility. The sensitivity as well as specificity was equal to 90 percent. It was not so sensitive (71 percent) in the category of infectious and parasitic diseases, such as HIV and diarrhoea, but its specificity was high at 97. The percentages were about the same for external cause of morbidity and mortality, (such as transport accidents, accidental drowning and submersion, and intentional self-harm) at 71 and 93 percent for sensitivity and specificity, respectively. The findings reveal that causes of death established by VA technique conform better to those established by medical diagnosis than do those the report of cause of death from death certificate.Percentage consistency to medically diagnosed causes of death to those established by VA was 66 percent while percentage consistency of causes of death from death certificate was 40 per cent. So the recommended solution in helping verify causes of death in Thailand and reduce numbers of unknown causes of death is to allow local health personnel with the VA method training course to identify and record causes of death where there is no medical diagnosis possible.
KEYWORD VERBAL AUTOPSY/ CAUSE OF DEATH /DEATH CERTIFICATE / SENSITIVITY/SPECIFICITY
 
FACULTY OF GRADUATE STUDIES. MAHIDOL UNIVERSITY. THAILAND
POWERED BY GITC.