RESUMO
The number of deaths caused by gastroenteritis rather than the notification of the illness is a more readily available index of the incidence of this disease in Jamaica. In order to examine the pattern of gastroenteritis with respect to its seasonal incidence data on a monthly or other cyclic basis. However, in the absence of such incidence figures for Jamaica, in this paper we have used the figures for deaths from gastroenteritis to determine the monthly incidence of disease in Jamaican children up to five years of age during the period 1963-1967(AU)
Assuntos
Gastroenterite/mortalidadeRESUMO
The exponential distribution in which the paramatic depends on one or several characteristics of patient or indices of severity of disease, are generally used for prognosis of survival time due to chronic diseases such as cancer. The assumption of such distribution of survival time is based on the fact that diagnosis of cancer in most of the cases is purely accidental in the sense that the patient might be suffering from the disease quite a long time before appearing for diagnosis or else an examination for a routine treatment may discover cancer. Therefore, the survival time is from time of diagnosis to death. Though in addition to the survival time, there are objective measures available which indicate the severity of disease at the time the patient first came under observation, yet in some studies of chronic disease, the survival time is the only quantitative response variable for statistical investigation. Under these circumstances the exponential distribution could still justified on the basis of its empirical fit with observed data which are available. This paper deals with the statistical theory of an exponential distribution which has found its most important application in blood cancer. In estimating the survival time for leukemia patients, the blast cell counts have been considered as the dependent variable or characteristic befitting the exponential distribution with two parameters. The parameters are estimated by the method of maximum likelihood and application has been shown with an example of sixteen leukemia patients who died with acute lymphoblastic, acute lymphatic, myeloid, monocytic and acute leukemia patients. In the example, the assumption of exponential distribution of survival time seemed justified by the way in which the results reproduced the original data. The comparison between the empirical survival distribution with the estimated survival curve associated with the median blast cells count, portrays the adequacy of the exponential distribution which is further tested with chi-square and found insignificant. It is striking that the survival probabilities depend strongly on blast cell counts. On a side investigation it was revealed that survival probabilities of male outweigh the female. The estimates of survival probabilities of individual patients with corresponding blast cell counts are also given in a monogram reckoning its important application from clinical trials point of view, when a patient's ability to respond to treatment depends more often on the severity of disease rather than effect of the treatment within the sample group. In conclusion, the statisitcal theory discussed in the paper has been further extended in its application in (1) prognosis of survival of leukemia patients with various treatments and therapy (2) the status of patient (dead or alive) at the end of study affecting treatment or therapy and (3) the evaluation of survival data from clinical trials (AU)
