392 Patient Reported Comorbidity Is Not A Reliable Data Source: A Systematic Review of The Literature

Introduction There is currently an increased demand for elective orthopaedic surgery. However, due to financial, time and resource limitations, there is a pressing need to identify those who would benefit most with the lowest risk of complications. Comorbidities play a fundamental part in this decision and the traditional way to ascertain this is through medical record data ion during pre-operative assessment, but this can be time-consuming and expensive. We therefore set out to establish whether patient self-reported comorbidities are reliable as a principal source of information. Method Searches of PubMed and Medline were performed by two independent researchers. Inclusion criteria were any published study assessing the reliability of at least one patient reported comorbidity against their medical record or clinical assessment as gold standard. Results There were 27 studies included 12 concluded with unreliability, 11 with reliability, 4 inconclusive. Factors found to affect the concordance included gender, age, ethnicity, level of education, living alone, marital status, number or severity of comorbidities and depression. Conclusions The majority of studies concluded that patient self-reported comorbidities are unreliable, even when their results showed good concordance. Although patient reported data is useful, it is not reliable enough to be used as a standalone measure.

Repeat immigration: A previously unobserved source of heterogeneity?

Aims: Register data allow for nuanced analyses of heterogeneities between sub-groups which are not observable in other data sources. One heterogeneity for which register data is particularly useful is in identifying unique migration histories of immigrant populations, a group of interest across disciplines. Years since migration is a commonly used measure of integration in studies seeking to understand the outcomes of immigrants. This study constructs detailed migration histories to test whether misclassified migrations may mask important heterogeneities. In doing so, we identify a previously understudied group of migrants called repeat immigrants, and show that they differ systematically from permanent immigrants. In addition, we quantify the degree to which migration information is misreported in the registers. Method: The analysis is carried out in two steps. First, we estimate income trajectories for repeat immigrants and permanent immigrants to understand the degree to which they differ. Second, we test data validity by cross-referencing migration information with changes in income to determine whether there are inconsistencies indicating misreporting. Results: From the first part of the analysis, the results indicate that repeat immigrants systematically differ from permanent immigrants in terms of income trajectories. Furthermore, income trajectories differ based on the way in which years since migration is calculated. The second part of the analysis suggests that misreported migration events, while present, are negligible. Conclusions: Repeat immigrants differ in terms of income trajectories, and may differ in terms of other outcomes as well. Furthermore, this study underlines that Swedish registers provide a reliable data source to analyze groups which are unidentifiable in other data sources.

Open-Source Data and the Study of Homicide

To date, no discussion has taken place in the social sciences as to the appropriateness of using open-source data to augment, or replace, official data sources in homicide research. The purpose of this article is to examine whether open-source data have the potential to be used as a valid and reliable data source in testing theory and studying homicide. Official and open-source homicide data were collected as a case study in a single jurisdiction over a 1-year period. The data sets were compared to determine whether open-sources could recreate the population of homicides and variable responses collected in official data. Open-source data were able to replicate the population of homicides identified in the official data. Also, for every variable measured, the open-sources captured as much, or more, of the information presented in the official data. Also, variables not available in official data, but potentially useful for testing theory, were identified in open-sources. The results of the case study show that open-source data are potentially as effective as official data in identifying individual- and situational-level characteristics, provide access to variables not found in official homicide data, and offer geographic data that can be used to link macro-level characteristics to homicide events.

Provincial Maternal Mortality Surveillance Systems in China

Background. Provincial maternal mortality surveillance systems (PMMSS) have been set up in nearly all the provinces in China to monitor local maternal mortality and provide the evidence for maternal health interventions suited to local conditions. However, till now little is known outside of China about the characteristics of PMMSS.Methods. A systematic review of the literature contained in PubMed and China Academic Journal Network Publishing database was carried out. The current situation on PMMSS was described. Provincial disparities on PMMR in six provinces were analyzed by Poisson regression analysis.Results. A total of 35 studies met the inclusion criteria, of which 31 were published in Chinese. PMMSS were set up and adjusted by the provincial government based on their own financial resources and demand. Provinces from remote region had the highest risk of maternal mortality, followed by provinces from inland region and coast region.Conclusions. PMMSS may be the most reliable data source for measuring provincial level MMR in each province. Great provincial disparities on PMMSS and PMMR do exist within the country; more emphasis should be placed on improving PMMSS and reducing PMMR particularly in the provinces with high maternal death burden.

Using Web Panels to Understand Whether Online Ad Exposure Influences Information-Seeking Behavior

Measuring the impact of online health campaigns is challenging. Traditionally, advertisement click-through rates (CTRs) have been used to measure message reach, but CTRs are low with most clicks coming from a small fraction of users. However, low CTRs do not necessarily indicate that an ad was not effective. There may be latency effects whereby people do not click on ads at time of exposure but visit the promoted website or conduct searches later. Online panels that unobtrusively collect panelists’ web behavior may provide a more reliable data source for measuring online campaign effects. We used web behavior data from a proprietary online panel to identify panelists who were either exposed or unexposed to the Tobacco Free Florida Cessation Internet ad campaign. We assessed whether ad exposure influenced website visits and searches on campaign-related topics up to 4 weeks after initial exposure. Those who were exposed to the campaign were significantly more likely than those who were not exposed to have visited the campaign website (0.65% vs. 0.13%, respectively, p < .001), but ad exposure did not influence searches on campaign-related topics. These results suggest that panel web behavior data may be useful for understanding behavioral response to and latency effects of online campaigns.

Late Silurian and Early Devonian Beyrichioidea from Gondwana and Perigondwanan terranes and their palaeobiogeographical implications

The Late Silurian to Early Devonian palaeogeography and the question whether there has been a large and deep Rheic Ocean between Laurussia and Gondwana is still a matter of discussion. It has been assumed in many papers that the Rheic Ocean formed a palaeogeographical barrier and, therefore, Gondwana had no common beyrichioidean ostracode faunas with Laurussia before the Emsian. The appearance of ostracodes (particularly the beyrichioideans) in Gondwana at this time should testify the closure of the Rheic Ocean. Although some palaeogeographic reconstructions consider common ostracodes in Laurussia and the Perigondwanan areas Armorica and Perunica in the Lochkovian, they do not refer to occurrences in Gondwana. However, several publications describe late Silurian to Pragian beyrichioideans from Gondwana and adjoined Terranes (Perigondwana), thus questioning the existence of the Rheic Ocean at this time. This paper aims to summarize the occurrences of common beyrichioidean ostracodes from (Peri-)Gondwana and Laurussia in an attempt to provide a reliable data source for future work. The occurrences of beyrichioideans from North Africa, South America, southeastern Turkey, Ibero-Armorica, and the Ossa Morena Zone (Spain) are reviewed and their ages are partially re-dated through correlation with other faunas. The presented results indicate that shallow marine migration paths must have existed, thus strengthening the arguments against the Rheic Ocean.