Soil erosion is a severe problem for many developing regions that lack adequate infrastructure to combat the problem. The authors established a first-order method for prioritizing areas to be examined and remediated using preexisting data and expert knowledge where data are lacking. The Universal Soil Loss Equation was applied to the Rio Lempa Basin in Central America using geographic information systems and remote sensing technologies, and the estimated erosion rates were compared with sediment delivery ratios. Spatial analysis indicates that agriculture on very steep slopes contributes only a small fraction to the total estimated soil erosion, whereas agriculture on gentle and moderately steep slopes contributes a large fraction of the erosion. Although much of the basin is in El Salvador, the greatest estimated amount of erosion is from Honduras. Data quality and availability were impaired by a lack of coordination among agencies and across countries. Several avenues for improving the authors' methods are described.

ObjectivesHandling missing information is an important methodological challenge in electronic health records research, with the reasons for missingness dictating how missing data should be handled. While there is extensive literature on methodological approaches to handling missing data, there is relatively little research with patients or clinicians investigating why information may be unrecorded in the medical records. A patient and public involvement event was held, embedded with a wider EHR event (UCL Festival of Digital Science -March 2016), to explore unrecorded information from a patient and researcher point of view with a view to developing further research in this area. ApproachCo-chaired by an academic and lay chair, members of the Farr Patient and Public Involvement Forum, academics, students and the general public attended workshop held during the Festival of Digital Health. A brief exercise was used to demonstrate the impact of missing data. Participants then worked in small facilitated groups to discuss situations where information about them (or people they knew) had not been recorded, the reasons why this information was not recorded and what the impact (or potential impact) was on the patient. A standardised data form was used by facilitators to record participants' conversations. The data forms were summarised into Excel and a brief thematic analysis was conducted. A written evaluation of the workshop was also completed by participants and included in the analysis. ResultsThe evaluation was overwhelming positive about embedding PPI workshop within an event with a wider remit, from both researchers and patient and public representatives. Preliminary findings on reasons for missingness included change of address or provider (both general practice and secondary care), individual clinician approaches to recording data, process of computerisation of medical notes, early stages of disease not reported to GP and unwillingness to disclose more sensitive information such as sexually transmitted disease. Discussion also included reasons for misrepresented or other errors with data. Further findings from the framework analysis will be presented. ConclusionThis exploratory workshop showed that patient and public involvement events can be embedded successfully within events aimed at professionals. Preliminary findings suggest that reasons for missingness often related to features of the provider or clinician as well as the patient.

No abstract