BackgroundDetection of Ureaplasma, Mycoplasma and Candida spp. in the vagina during pregnancy has previously been associated with preterm birth (PTB). However, the prevalence of these microorganisms and the associated obstetric risks (likely to be population-specific) have not been determined in Australian women; furthermore, in the case of Ureaplasma spp., very few studies have attempted characterisation at the species level and none have examined genotype/serovar status to further refine risk assessment.MethodsIn order to address these issues we sampled the vaginal fluid of 191 pregnant Australian women at three time points in pregnancy. Culture methods were used for detection of Ureaplasma spp. and Candida spp., and real-time PCR was used for speciation of U. parvum and U. urealyticum, non-albicans Candida spp., Mycoplasma hominis and Mycoplasma genitalium. High-resolution melt PCR was used to genotype U. parvum. Data on various lifestyle factors (including sex during pregnancy and smoking), antimicrobial use and pregnancy outcome were collected on all participants. Chi-square tests were used to assess the association of vaginal microorganisms with PTB.ResultsDetection of Ureaplasma spp. was higher among spontaneous PTB cases, specifically in the presence of U. parvum [77 % preterm (95 % confidence interval (CI) 50–100 %) vs. 36 % term (CI: 29–43 %), p = 0.004], but not U. urealyticum. The association with PTB strengthened when U. parvum genotype SV6 was detected (54 % preterm (CI: 22–85 %) vs. 15 % term (CI: 10–20 %), p = 0.002); this genotype was also present in 80 % (4/5) of cases of PTB <34 weeks gestation. When present with Candida albicans in the same sample, the association with PTB remained strong for both U. parvum [46 % preterm (CI: 15–78 %) vs. 13 % term (CI: 8–18 %), p = 0.005] and U. parvum genotype SV6 [39 % preterm (CI: 8–69 %) vs. 7 % term (CI: 3–11 %), p = 0.003]. With the exception of Candida glabrata, vaginal colonisation status for all organisms was stable throughout pregnancy. Smoking significantly increased the likelihood of detection of all target organisms.ConclusionsThese data suggest that the presence of different species and serovars of Ureaplasma spp. in the vagina confers an increased risk of spontaneous PTB, findings which may be useful in risk assessment for identifying women who would benefit from antimicrobial treatment.

It is useful for reviewers of economic evaluations to assess quality in a manner that is consistent and comprehensive. Checklists can allow this, but there are concerns about their reliability and how they are used in practice. We aimed to describe how checklists have been used in systematic reviews of health economic evaluations. Methods: Meta-review with snowball sampling. We compiled a list of checklists for health economic evaluations and searched for the checklists' use in systematic reviews from January 2010 to February 2018. We extracted data regarding checklists used, stated checklist function, subject area, number of reviewers, and issues expressed about checklists. Results: We found 346 systematic reviews since 2010 that used checklists to assess economic evaluations. The most common checklist in use was developed in 1996 by Drummond and Jefferson, and the most common stated use of a checklist was quality assessment. Checklists and their use varied within subject areas; 223 reviews had more than one reviewer who used the checklist. Conclusions: Use of checklists is inconsistent. Eighteen individual checklists have been used since 2010, many of which have been used in ways different from those originally intended, often without justification. Different systematic reviews in the same subject areas would benefit from using one checklist exclusively, using checklists as intended, and having 2 reviewers complete the checklist. This would increase the likelihood that results are transparent and comparable over time.

No abstract

BackgroundEstimating health care costs, either in the context of understanding resource utilization in the implementation of a health plan, or in the context of economic evaluation, has become a common activity of health planners, health technology assessment agencies and academic groups. However, data sources for costs outside of direct service delivery are often scarce. WHO-CHOICE produces global price databases and guidance on quantity assumptions to support country level costing exercises. This paper presents updates to the WHO-CHOICE methodology and price databases for programme costs.MethodsWe collated publicly available databases for 14 non-traded cost variables, as well as a set of traded items used within health systems (traded goods are those which can be purchased from anywhere in the world, whereas non-traded goods are those which must be produced locally, such as human resources). Within each of the variables, missing data was present for some proportion of the WHO member states. For each variables statistical or econometric models were used to model prices for each of the 194 WHO member states in 2010 International Dollars. Literature reviews were used to update quantity assumptions associated with each variable to contribute to the support costs of disease control programmes.ResultsA full database of prices for disease control programme support costs is available for country-specific costing purposes. Human resources are the largest driver of disease control programme support costs, followed by supervision costs.ConclusionsDespite major advances in the availability of data since the previous version of this work, there are still some limitations in data availability to respond to the needs of those wishing to develop cost and cost-effectiveness estimates. Greater attention to programme support costs in cost data collection activities would contribute to an understanding of how these costs contribute to quality of health service delivery and should be encouraged.