| Objective:A quality analysis was conducted to examine the level of quality of filling the cause-of-death chain and selecting underlying causes of death for73270medical certificate of cause of death during2011-2012in a city. The Specificity and sensitivity were used to assess the characteristic and extent of misclassification of cause of death in cause-of-death spectrum, thus providing a reference for improving the quality of vital registration system of china.Methods:Filling the cause-of-death chain and selecting underlying causes of death as the two major components of completing medical certificate of cause of death is the core of quality assessment in this study. Related researches about the quality assessment of medical certificate of cause of death were used to establish the index system of this study. The assessment index system of filling the cause-of-death chain include six indicators:①just filling out the clinical signs and symptoms;②haven’t reporting more specific disease;③General information;④unknown Cause of death;⑤Filling inappropriate content related injuries⑥Order of chaos. The assessment index system of selecting underlying causes of death include nine indicators:①Ignoring the use of the general principles;②Ignoring the use of Rule1;③Ignoring the use of Rule2;④Ignoring the use of Rule3;⑤Ignoring the use of Rule A;⑥Ignoring the use of Rule B;⑦Ignoring the use of Rule C;⑧Ignoring the use of Rule D;⑨Ignoring the use of Rule E;⑩Ignoring the use of Rule E.The indicators above were used to quality assessment of73270medical certificate of cause of death; meanwhile a new underlying cause of death and ICD-10Codes were given when the medical certificate of cause of death was proved be wrong in the selecting underlying causes of death process. A analysis was applied to describe the distribution of age, Highest diagnostic unit type, Reporting units type, and the joint distribution of the two major filling errors. The way of Classification of Diseases in ICD-10was used to describe the characteristic and pattern of the51major cause of death. Sensitivity and positive predictive value were used to evaluate the cause of death misclassification. All statistical analyses were performed in SPSS18.0software package and Excel2010.Result:1. Basic conditions of investigation objectThis study collected74520medical certificate of cause of death. Through the screening, we finally determined73270medical certificate of cause of death to meet the requirements. A total of2011collection of qualified medical certificate is36038and36332copies in2012. The number of men is44516; the number of women is28754.Death site is divided into six classes:the hospital ward, emergency room, the way to the hospital, home, other and unknown, its constituent ratio was22.64%,4.64%,4.10%,63.50%,2.58%,2.53%.The highest diagnostic unit is divided into seven classes:provincial hospitals, municipal hospitals, County-level hospitals, Community Health Center, township hospitals, no treatment, and the other or unknown; The constituent ratio is27.36%,22.41%,25.37%,11.50%,4.35%,6.33%,2.69%.Reporting unit is divided into five categories:general hospital, specialized hospitals, community health service centers, township hospitals and other, the proportion is20.92%,4.32%,14.30%,59.06%,1.41%.2. Quality assessmentThe error rate of filling cause-of-death chain in this study is21.62%, which the six type of error rate about just filling out the clinical signs and symptoms, haven’t reporting more specific disease, General information, unknown Cause of death, Filling inappropriate content related injuries, is2.29%,8.24%,4.34%,3.84%,2.13%,0.80%.The error rate of selecting underlying cause of death is26.46%, the main error types are:(1) the general principle of use;(2) ignore use rules of c. contact, the two types of errors of the underlying cause inference error in the study of the majority (99.68%). The right proportion of the two main processes, filling cause-of-death chain and selecting underlying cause of death, is68.46%. The wrong proportion of the two main processes, filling cause-of-death chain and selecting underlying cause of death, is16.53%. The errors ratio of a single process is15.02%, the single error ratio of filling cause-of-death chain is5.09%, and the single error ratio of selecting underlying cause of death is9.93%.3. The spectrum and misclassification of cause of deathBefore and after the adjustment, the "circulatory system disease","cancer","the respiratory system disease","injury""has always been the top four cause of death on the cause of death spectrum, and the order haven’t change.Before the adjustment,"circulatory system disease","cancer","the respiratory system disease","injury" cases are:37068,13645,6990,13645; After adjusting the four cases are:35462,13142,7193,13142; the number of cases have small variations before and after adjustment. The order of other causes of death on the cause of death spectrum change slightly. The order of "symptoms, signs and clinical and laboratory abnormalities can see, not classified elsewhere" rise from the sixth to the fifth, the number of cases rise from2061to4265. The order of "Other cause of death" rise from fifth to sixth, the number of cases rises from3479to3536. The order of "The digestive system disease" dropped from8to7, the number of cases dropped from1686to1621. The order of "Diabetes" rise from seventh to eighth, but the number of cases dropped from1588to1546.All kinds of causes of death are present misclassification. Sensitivity and positive predictive value was used to evaluate this problem. The cause of death such as "infectious diseases and parasitic diseases","cancer","hurting" have the sensitivity and positive predictive value both of more than80%,"Blood and hematopoietic organs disease and involved in the immune mechanism of some diseases","diabetes","the digestive system disease" have the sensitivity and positive predictive value between70%and80%."diabetes","circulatory system disease"," respiratory disease","the symptoms, signs and clinical and laboratory abnormalities can see, not classified elsewhere" have sensitivity and positive predictive value of much lower than50%.Conclusion:1. The high error rate of filling the cause-of-death chain and selecting underlying causes of death are the two main problems in the vital registration system.2. The average number of stages of cause-of-death chain in medical certificate of cause of death is too low, which seriously affecting the process of selecting underlying cause of death. On the other hand, it also Indicate that the level of vital registration system needs to be improved.3. Type of reporting unit will affect the reporting quality of medical certificate of cause of death, the reporting quality of township hospitals and community health center is less than the general hospitals and specialized hospitals.4. The compensation effect of misclassification in some cause of death is very uneven, such as "Idiopathic (primary) hypertension","symptoms, signs and abnormal clinical and laboratory findings, not elsewhere classified";"the other injuries","other chronic","Other hypertension","lower respiratory diseases","subarachnoid hemorrhage". |
PDF Full Text Request