Differences in reported sepsis incidence according to study design: a literature review.

BACKGROUND: Sepsis and severe sepsis are common conditions in hospital settings, and are associated with high rates of morbidity and mortality, but reported incidences vary considerably. In this literature review, we describe the variation in reported population-based incidences of sepsis and severe sepsis. We also examine methodological and demographic differences between studies that may explain this variation. METHODS: We carried out a literature review searching three major databases and reference lists of relevant articles, to identify all original studies reporting the incidence of sepsis or severe sepsis in the general population. Two authors independently assessed all articles, and the final decision to exclude an article was reached by consensus. We extracted data according to predetermined variables, including study country, sepsis definition, and data source. We then calculated descriptive statistics for the reported incidences of sepsis and severe sepsis. The studies were classified according to the method used to identify cases of sepsis or severe sepsis: chart-based (i.e. review of patient charts) or code-based (i.e. predetermined International Classification of Diseases [ICD] codes). RESULTS: Among 482 articles initially screened, we identified 23 primary publications reporting incidence of sepsis and/or severe sepsis in the general population. The reported incidences ranged from 74 to 1180 per 100,000 person-years and 3 to 1074 per 100,000 person-years for sepsis and severe sepsis, respectively. Most chart-based studies used the Bone criteria (or a modification hereof) and Protein C Worldwide Evaluation in Severe Sepsis (PROWESS) study criteria to identify cases of sepsis and severe sepsis. Most code-based studies used ICD-9 codes, but the number of codes used ranged from 1 to more than 1200. We found that the incidence varied according to how sepsis was identified (chart-based vs. code-based), calendar year, data source, and world region. CONCLUSION: The reported incidences of sepsis and severe sepsis in the general population varied greatly between studies. Such differences may be attributable to differences in the methods used to collect the data, the study period, or the world region where the study was undertaken. This finding highlights the importance of standardised definitions and acquisition of data regarding sepsis and severe sepsis.

Maternal Pre-pregnancy BMI and Reproductive Health of Daughters in Young Adulthood.

Objective To investigate the possible associations between maternal pre-pregnancy body mass index (BMI) and daughters' age of menarche and subsequent markers of reproductive health. Methods Nine hundred eighty-five pregnant women (80 %) were enrolled at their routine 30th week examinations in 1988-1989. In 2008, a follow-up questionnaire was completed for 365 daughters (83 %), while 267 daughters (61 %) participated in a subsequent clinical examination. Main outcome measures were age of menarche, reproductive hormone profile, and ovarian follicle count in daughters. Results Daughters of mothers in the highest pre-pregnancy BMI tertile (BMI ≥ 22.0 kg/m(2)) had an adjusted 4.1 (0.3; 8.0) months earlier menarche compared with the middle tertile group (BMI 20.0-21.9 kg/m(2)). Among non-users of hormonal contraceptives, daughters of mothers in the highest pre-pregnancy BMI tertile had non-significantly lower dehydroepiandrosterone-sulphate (DHEAS), estradiol, and free estrogen index (FEI), compared to the middle BMI tertile. This was supported by a sub-analysis using the WHO classification (underweight, BMI < 18.50; normal range, BMI 18.50-24.99; overweight/obese, BMI ≥ 25.00 kg/m(2)) as exposure groups, in which daughters of overweight mothers had lower levels of DHEAS and estradiol, and lower FEI compared to daughters of normal weight mothers. No associations were found for ovarian follicle count in any of the groups. Conclusions for Practice We found that higher maternal BMI is associated with earlier age of menarche in daughters. A possible impact of maternal pre-pregnancy BMI on DHEAS and estradiol serum levels, and FEI in non-users of hormonal contraceptives was indicated, but the results were not statistically significant.