Article

Confidence intervals for directly standardized rates: a method based on the gamma distribution.

National Cancer Institute, Division of Cancer Prevention and Control, Bethesda, Maryland 20892-7354, USA.
Statistics in Medicine (Impact Factor: 2.04). 05/1997; 16(7):791-801. DOI: 10.1002/(SICI)1097-0258(19970415)16:73.0.CO;2-#
Source: PubMed

ABSTRACT We offer an approximation to central confidence intervals for directly standardized rates, where we assume that the rates are distributed as a weighted sum of independent Poisson random variables. Like a recent method proposed by Dobson, Kuulasmaa, Eberle and Scherer, our method gives exact intervals whenever the standard population is proportional to the study population. In cases where the two populations differ non-proportionally, we show through simulation that our method is conservative while other methods (the Dobson et al. method and the approximate bootstrap confidence method) can be liberal.

Full-text

Available from: Michael P Fay, Aug 18, 2014
0 Followers
 · 
147 Views
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background Tick-borne diseases are a major health risk for humans and dogs. In addition to collection and analysis of questing ticks, analysis of host-associated ticks for the presence of pathogens is a valuable method to gain insight into transmission patterns of tick-borne diseases.Methods Ticks were collected from dogs living in the Berlin/Brandenburg area. The three tick species Ixodes ricinus, Ixodes hexagonus and Dermacentor reticulatus were examined for the presence of Babesia spp., Borrelia spp., Rickettsia spp. and Anaplasmataceae. Conventional PCR followed by sequencing was used for pathogen detection and characterization.Results Babesia spp. were found in 2.5% and 3% of I. ricinus and I. hexagonus, respectively. Sequencing revealed the presence of Babesia microti, Babesia capreoli and Babesia venatorum. D. reticulatus were free of Babesia canis. Rickettsia spp. were detected in 61% of I. ricinus, 44% of I. hexagonus and 39% of D. reticulatus. Specifically detected were Rickettsia raoulti in D. reticulatus and I. hexagonus, Rickettsia helvetica in I. ricinus and I. hexagonus and Rickettsia monacensis in I. hexagonus. Anaplasma phagocytophilum and Candidatus Neoehrlichia mikurensis have been reported previously in I. ricinus (6.5% and 4.3%, respectively) and I. hexagonus (3.9% and 5.9%). Borrelia spp. were found in 11.6% of I. ricinus and 11.2% of I. hexagonus. Subsequent genospecies analysis revealed Borrelia afzelii, Borrelia garinii, Borrelia burgdorferi sensu stricto and Borrelia miyamotoi. Simultanous presence of more than one pathogen was found in 20% of I. ricinus and in 59% of I. hexagonus whereas the total frequency of any pathogen was 65% in I. ricinus, 59% in I. hexagonus and 64% in D. reticulatus. Ticks in which A. phagocytophilum was detected had a significantly increased risk of also containing Rickettsia. Ticks harbouring a pathogen had significantly higher scutal indices than ticks without presence of any pathogen.Conclusions Frequencies of potential human or canine pathogens in ticks were considerable and DNA of all four groups of pathogens was detected. Differences in scutal indices might suggest that pathogens are frequently taken up by ticks when feeding on dogs in Berlin/Brandenburg.
    Parasites & Vectors 12/2014; 7(1):535. DOI:10.1186/PREACCEPT-1653925552127672 · 3.25 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Objectives: Poison control centres provide information on the management of poisoning incidents. The British Columbia (BC) Drug and Poison Information Centre recently implemented an electronic database system for recording case information, making it easier to use case data as a potential source of population-based information on health services usage and health status. This descriptive analysis maps poisoning case rates in BC, highlighting differences in patient age, substance type, medical outcome, and caller location. Methods: There were 50,621 human exposure cases recorded during 2012 and 2013. Postal code or city name was used to assign each case to a Health Service Delivery Area (HSDA). Case rates per 1,000 person-years were calculated, including crude rates, age-standardized rates, age-specific rates, and rates by substance type, medical outcome, and caller location.
    10/2014; 6(2). DOI:10.5210/ojphi.v6i2.5376
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Background: Previous studies have reported a high incidence of amyotrophic lateral sclerosis (ALS) in endemic foci in the Kii Peninsula, Japan. However, little is known about the ALS frequency in the whole country. Furthermore, the presence of ethnic variation in the incidence of ALS remains unknown.Methods: We conducted a nationwide survey of ALS frequency in 2013 to estimate its annual prevalence and incidence. ALS was diagnosed based on the El Escorial Criteria. The study period was the 2009 fiscal year, from April 2009 to March 2010. To compare the incidence of ALS among prefectures, standardized incidence ratios (SIRs) and 95% confidence intervals (CIs) were calculated under the assumption of Poisson distribution.Results: The annual crude prevalence and incidence rates per 100 000 people per year were 9.9 (95% CI 9.7-10.1) and 2.2 (95% CI 2.1-2.3), respectively. The age group with the highest prevalence as well as incidence was 70-79 years, and the male-female ratio was approximately 1.5. The annual incidence rate adjusted for age and sex using the 2000 U.S. standard population was 2.3 (95% CI 2.2-2.4) per 100 000 people. Some prefectures had significantly high SIRs: Okinawa, Nara and Wakayama in the Kii Peninsula, and Niigata for males; Kumamoto for females.Conclusions: This is the first report on the annual prevalence and incidence of ALS in the representative population of Japan. We identified some prefectures with a high incidence of ALS. However, the incidence of ALS in the Japanese population was much lower than in the Caucasian populations of Europe and North America.
    Journal of Epidemiology 08/2014; 24(6). DOI:10.2188/jea.JE20140059 · 2.86 Impact Factor