Machine learning-based prediction of medication refill adherence among first-time insulin users with type 2 diabetes.

AIMS: The prevalence of Type 2 Diabetes Mellitus (T2DM) is projected to be 7 % in 2030. Despite its need for long-term diabetes care, the adherence rate of injectable medications such as insulin is around 60 %, lower than the acceptable threshold of 80 %. This study aims to create classification models to predict insulin adherence among adult T2DM naïve insulin users. METHODS: Clinical data were extracted from Taipei Medical University Clinical Research Database (TMUCRD) from January 1st, 2004 to December 30th, 2020. A patient was regarded as adherent if his/her medication possession ratio (MPR) was at least 80 %. Seven domains of predictors were created, including demographics, baseline medications, baseline comorbidities, baseline laboratory data, healthcare resource utilization, index insulins, and the concomitant non-insulin T2DM medications. We built two Xgboost models for internal and external testing respectively. RESULTS: Using a cohort of 4134 patients from Taiwan, our model achieved the Area Under the curve of the Receiver Operating Characteristic (AUROC) of the internal test was 0.782 and the AUROC of the external test was 0.771. the SHAP (SHapley Additive exPlanations) value showed that the number of prescribed medications, the number of outpatient visits, and laboratory data were predictive of future insulin adherence. CONCLUSIONS: This is the first study to predict adherence among adult naïve insulin users. The developed model is a potential clinical decision support tool to identify possible non-adherent patients for healthcare providers to design individualized education plans.

Assessing the Job Satisfaction of Registered Nurses Using Sentiment Analysis and Clustering Analysis.

Job satisfaction is a critical component of the professional work environment and is often ascertained through surveys that include structured or open-ended questions. Using data from 24,543 respondents to California Board of Registered Nursing biennial surveys, this study examines the job satisfaction of registered nurses (RNs) by applying clustering analysis to structured job satisfaction items and sentiment analysis to free-text comments. The clustering analysis identified three job satisfaction groups (low, medium, and high satisfaction). Sentiment analysis scores were significantly associated with the job satisfaction groups in both bivariate and multivariate analyses. Differences between the job satisfaction clusters were mostly driven by satisfaction with workload, adequacy of the clerical support services, adequacy of the number of RN staff, and skills of RN colleagues. In addition, there was dispersion in satisfaction related to involvement in management and policy decisions, recognition for a job well done, and opportunities for professional development.

Extracting Eligibility Criteria from the Narrative Text of Scientific Research Articles.

Eligibility criteria among hundreds of National Health Insurance Research Database (NHIRD) research papers have similar constituent elements, such as demographic characteristics or diagnostic codes. The study results of the same disease could vary among different research due to the variation of the criteria statements, therefore the narrative patterns analysis tool would be helpful for summarizing the knowledge implicitly contained in the eligibility criteria. In this study, we developed a series of R-based text processing methods to extract the narrative eligibility criteria in NHIRD papers by simplifying the article titles and content paragraphs, identifying medical concepts and abbreviations, then detecting basic demographic characteristics and ICD-9-CM diagnosis codes. Although there is still room for improvement on study type identifying, the high performance in classifying the study type, detecting age restrictions and extracting ICD-9-CM codes still shows the system usefulness for the analysis of eligibility criteria.

Building Chronic Kidney Disease Clinical Practice Guidelines Using the openEHR Guideline Definition Language.

BACKGROUND: As a result of the disease's high prevalence, chronic kidney disease (CKD) has become a global public health problem. A clinical decision support system that integrates with computer-interpretable guidelines (CIGs) should improve clinical outcomes and help to ensure patient safety. OBJECTIVES: The openEHR guideline definition language (GDL) is a formal language used to represent CIGs. This study explores the feasibility of using a GDL approach for CKD; it also attempts to identify any potential gaps between the ideal concept and reality. METHODS: Using the Kidney Disease Improving Global Outcomes (KDIGO) anemia guideline as material, we designed a development workflow in order to establish a series of GDL guidelines. Focus group discussions were conducted in order to identify important issues related to GDL implementation. RESULTS: Ten GDL guidelines and 37 archetypes were established using the KDIGO guideline document. For the focus group discussions, 16 clinicians and 22 IT experts were recruited and their perceptions, opinions and attitudes towards the GDL approach were explored. Both groups provided positive feedback regarding the GDL approach, but raised various concerns about GDL implementation. CONCLUSIONS: Based on the findings of this study, we identified some potential gaps that might exist during implementation between the GDL concept and reality. Three directions remain to be investigated in the future. Two of them are related to the openEHR GDL approach. Firstly, there is a need for the editing tool to be made more sophisticated. Secondly, there needs to be integration of the present approach into non openEHR-based hospital information systems. The last direction focuses on the applicability of guidelines and involves developing a method to resolve any conflicts that occur with insurance payment regulations.

An exploratory study using an openEHR 2-level modeling approach to represent common data elements.

BACKGROUND AND OBJECTIVE: In order to facilitate clinical research across multiple institutions, data harmonization is a critical requirement. Common data elements (CDEs) collect data uniformly, allowing data interoperability between research studies. However, structural limitations have hindered the application of CDEs. An advanced modeling structure is needed to rectify such limitations. The openEHR 2-level modeling approach has been widely implemented in the medical informatics domain. The aim of our study is to explore the feasibility of applying an openEHR approach to model the CDE concept. MATERIALS AND METHODS: Using the National Institute of Neurological Disorders and Stroke General CDEs as material, we developed a semiautomatic mapping tool to assist domain experts mapping CDEs to existing openEHR archetypes in order to evaluate their coverage and to allow further analysis. In addition, we modeled a set of CDEs using the openEHR approach to evaluate the ability of archetypes to structurally represent any type of CDE content. RESULTS: Among 184 CDEs, 28% (51) of the archetypes could be directly used to represent CDEs, while 53% (98) of the archetypes required further development (extension or specialization). A comprehensive comparison between CDEs and openEHR archetypes was conducted based on the lessons learnt from the practical modeling. DISCUSSION: CDEs and archetypes have dissimilar modeling approaches, but the data structure of both models are essentially similar. This study proposes to develop a comprehensive structure to model CDE concepts instead of improving the structure of CED. CONCLUSION: The findings from this research show that the openEHR archetype has structural coverage for the CDEs, namely the openEHR archetype is able to represent the CDEs and meet the functional expectations of the CDEs. This work can be used as a reference when improving CDE structure using an advanced modeling approach.

The effect of particulate matter size on cardiovascular health in Taipei Basin, Taiwan.

BACKGROUND: Although the overall effect of particulate matter (PM) on cardiovascular disease (CVD) has been previously documented, the effect of different PM sizes (PM10, PM2.5-10 and PM2.5) has not been well studied. This study estimates the effect of different PM sizes on the incidence of CVD in Taipei, Taiwan. METHODS: We collected outpatients with CVD from 2006 to 2010 and data on the concentrations of air pollutants such as PM10, PM2.5-10, PM2.5, sulfur dioxide, carbon monoxide, nitrogen dioxide, and ozone. A Distributed Lag Non-linear Model (DLNM) was used to explore the effect of different PM sizes on CVD risk. RESULTS: In high air pollution events, PM2.5 was significantly associated with elevated risk (4.9%) [95% confidence interval (CI): 1.010-1.089] for CVD with increasing interquartile range (IQR) in single air pollutant model. PM2.5-10 and PM10 did not show a significant positive association with CVD in this study. After adjusting for other air pollutants such as SO2, CO, NO2, and O3, the estimated effect of PM2.5 only decreased 0.2%. Moreover, patients under 40 years old did not show a significant association between PM2.5 and CVD. CONCLUSION: This study demonstrates that only PM2.5 is significantly positively correlated with the number of daily outpatient visits for CVD during high air pollution events.

A rational approach to estimating the surgical demand elasticity needed to guide manpower reallocation during contagious outbreaks.

BACKGROUND: Emerging infectious diseases continue to pose serious threats to global public health. So far, however, few published study has addressed the need for manpower reallocation needed in hospitals when such a serious contagious outbreak occurs. AIM: To quantify the demand elasticity of the major surgery types in order to guide future manpower reallocation during contagious outbreaks. MATERIALS AND METHODS: Based on a nationwide research database in Taiwan, we extracted the monthly volumes of major surgery types for the period 1998-2003, which covered the SARS period, in order to carry out a time series analysis. The demand elasticity of each surgery type was then estimated by autoregressive integrated moving average (ARIMA) analysis. RESULTS: During the study period, the surgical volumes of most selected surgery types either increased or remained steady. We categorized these surgery types into low-, moderate- and high-elastic groups according to their demand elasticity. Appendectomy, 'open reduction of fracture with internal fixation' and 'free skin graft' were in the low demand elasticity group. Transurethral prostatectomy and extracorporeal shockwave lithotripsy (ESWL) were in the high demand elasticity group. The manpower of the departments carrying out the surgeries with low demand elasticity should be maintained during outbreaks. In contrast, departments in charge of surgeries mainly with high demand elasticity, like urology departments, may be in a position to have part of their staff reallocated. CONCLUSIONS: Taking advantage of the demand variation during the SARS period in 2003, we adopted the concept of demand elasticity and used a time series approach to figure out an effective index of demand elasticity for various types of surgery that could be used as a rational reference to carry out manpower reallocation during contagious outbreak situations.

Applying data mining for the analysis of breast cancer data.

Data mining, also known as Knowledge-Discovery in Databases (KDD), is the process of automatically searching large volumes of data for patterns. For instance, a clinical pattern might indicate a female who have diabetes or hypertension are easier suffered from stroke for 5 years in a future. Then, a physician can learn valuable knowledge from the data mining processes. Here, we present a study focused on the investigation of the application of artificial intelligence and data mining techniques to the prediction models of breast cancer. The artificial neural network, decision tree, logistic regression, and genetic algorithm were used for the comparative studies and the accuracy and positive predictive value of each algorithm were used as the evaluation indicators. 699 records acquired from the breast cancer patients at the University of Wisconsin, nine predictor variables, and one outcome variable were incorporated for the data analysis followed by the tenfold cross-validation. The results revealed that the accuracies of logistic regression model were 0.9434 (sensitivity 0.9716 and specificity 0.9482), the decision tree model 0.9434 (sensitivity 0.9615, specificity 0.9105), the neural network model 0.9502 (sensitivity 0.9628, specificity 0.9273), and the genetic algorithm model 0.9878 (sensitivity 1, specificity 0.9802). The accuracy of the genetic algorithm was significantly higher than the average predicted accuracy of 0.9612. The predicted outcome of the logistic regression model was higher than that of the neural network model but no significant difference was observed. The average predicted accuracy of the decision tree model was 0.9435 which was the lowest of all four predictive models. The standard deviation of the tenfold cross-validation was rather unreliable. This study indicated that the genetic algorithm model yielded better results than other data mining models for the analysis of the data of breast cancer patients in terms of the overall accuracy of the patient classification, the expression and complexity of the classification rule. The results showed that the genetic algorithm described in the present study was able to produce accurate results in the classification of breast cancer data and the classification rule identified was more acceptable and comprehensible.

Comparison of a semi-automatic annotation tool and a natural language processing application for the generation of clinical statement entries.

BACKGROUND AND OBJECTIVE: Electronic medical records with encoded entries should enhance the semantic interoperability of document exchange. However, it remains a challenge to encode the narrative concept and to transform the coded concepts into a standard entry-level document. This study aimed to use a novel approach for the generation of entry-level interoperable clinical documents. METHODS: Using HL7 clinical document architecture (CDA) as the example, we developed three pipelines to generate entry-level CDA documents. The first approach was a semi-automatic annotation pipeline (SAAP), the second was a natural language processing (NLP) pipeline, and the third merged the above two pipelines. We randomly selected 50 test documents from the i2b2 corpora to evaluate the performance of the three pipelines. RESULTS: The 50 randomly selected test documents contained 9365 words, including 588 Observation terms and 123 Procedure terms. For the Observation terms, the merged pipeline had a significantly higher F-measure than the NLP pipeline (0.89 vs 0.80, p<0.0001), but a similar F-measure to that of the SAAP (0.89 vs 0.87). For the Procedure terms, the F-measure was not significantly different among the three pipelines. CONCLUSIONS: The combination of a semi-automatic annotation approach and the NLP application seems to be a solution for generating entry-level interoperable clinical documents.

A multi-technique approach to bridge electronic case report form design and data standard adoption.

BACKGROUND AND OBJECTIVE: The importance of data standards when integrating clinical research data has been recognized. The common data element (CDE) is a consensus-based data element for data harmonization and sharing between clinical researchers, it can support data standards adoption and mapping. However, the lack of a suitable methodology has become a barrier to data standard adoption. Our aim was to demonstrate an approach that allowed clinical researchers to design electronic case report forms (eCRFs) that complied with the data standard. METHODS: We used a multi-technique approach, including information retrieval, natural language processing and an ontology-based knowledgebase to facilitate data standard adoption using the eCRF design. The approach took research questions as query texts with the aim of retrieving and associating relevant CDEs with the research questions. RESULTS: The approach was implemented using a CDE-based eCRF builder, which was evaluated using CDE- related questions from CRFs used in the Parkinson Disease Biomarker Program, as well as CDE-unrelated questions from a technique support website. Our approach had a precision of 0.84, a recall of 0.80, a F-measure of 0.82 and an error of 0.31. Using the 303 testing CDE-related questions, our approach responded and provided suggested CDEs for 88.8% (269/303) of the study questions with a 90.3% accuracy (243/269). The reason for any missed and failed responses was also analyzed. CONCLUSION: This study demonstrates an approach that helps to cross the barrier that inhibits data standard adoption in eCRF building and our evaluation reveals the approach has satisfactory performance. Our CDE-based form builder provides an alternative perspective regarding data standard compliant eCRF design.

Noncancer hospice care in taiwan: a nationwide dataset analysis from 2005 to 2010.

UNLABELLED: Abstract Background: The National Health Insurance program (NHI) in Taiwan has provided hospice services since 2000, and it was expanded to noncancer illnesses in September 2009. The issues of noncancer hospice care and the impact of the expanded hospice policy remain unclear. METHODS: Data were collected retrospectively from claims data of hospice admissions using the NHI Research Database of 2005-2010. RESULTS: A total of 359 noncancer subjects and 412 hospice admissions were enrolled; 1795 age- and gender-matched cancer patients and 2578 hospice admissions were selected as a comparison group. Noncancer hospice care increased markedly after the third quartile of 2009. The most common noncancer diagnosis was "other diseases of the lung" (23.9%). The noncancer subjects had a significantly lower frequency of admissions, lower Charlson Comorbidity Index (CCI) scores, shorter hospice stay, and higher mortality rate than the cancer subjects. Family physicians provided the majority of hospice services in both groups. Acute low respiratory conditions (ALRC) were the most common acute comorbidity in deceased subjects. The noncancer decedents had more ALRC, sepsis/bacteremia, nontraumatic shock, acute myocardial infarctions, and esophageal varicose bleeding than the comparison group. The mean inpatient charges differed insignificantly between both groups. The noncancer subjects correlated negatively with CCI (odds ratio [OR] 0.59 in all hospice admissions; 0.63 in decedents), but positively with a hospice stay ≤3 days, mortality, sepsis/bacteremia, ALRC, nontraumatic shock, and acute myocardial infarctions compared with the cancer subjects (OR 1.42, 1.98, 2.24, 2.36, 2.17, and 11.68, respectively, adjusted by CCI). CONCLUSIONS: The expanded palliative care policy has impacted positively on noncancer hospice care in Taiwan. The terminal noncancer patients had higher risks for short hospice stay, sepsis, nontraumatic shock, and respiratory and heart problems than the cancer subjects. Early referral to hospices is required for terminal patients in Taiwan. The CCI had a limited role for cost/severity evaluations of hospice care.

A multi-part matching strategy for mapping LOINC with laboratory terminologies.

OBJECTIVE: To address the problem of mapping local laboratory terminologies to Logical Observation Identifiers Names and Codes (LOINC). To study different ontology matching algorithms and investigate how the probability of term combinations in LOINC helps to increase match quality and reduce manual effort. MATERIALS AND METHODS: We proposed two matching strategies: full name and multi-part. The multi-part approach also considers the occurrence probability of combined concept parts. It can further recommend possible combinations of concept parts to allow more local terms to be mapped. Three real-world laboratory databases from Taiwanese hospitals were used to validate the proposed strategies with respect to different quality measures and execution run time. A comparison with the commonly used tool, Regenstrief LOINC Mapping Assistant (RELMA) Lab Auto Mapper (LAM), was also carried out. RESULTS: The new multi-part strategy yields the best match quality, with F-measure values between 89% and 96%. It can automatically match 70-85% of the laboratory terminologies to LOINC. The recommendation step can further propose mapping to (proposed) LOINC concepts for 9-20% of the local terminology concepts. On average, 91% of the local terminology concepts can be correctly mapped to existing or newly proposed LOINC concepts. CONCLUSIONS: The mapping quality of the multi-part strategy is significantly better than that of LAM. It enables domain experts to perform LOINC matching with little manual work. The probability of term combinations proved to be a valuable strategy for increasing the quality of match results, providing recommendations for proposed LOINC conepts, and decreasing the run time for match processing.

A longitudinal cohort study of incidence rates of inguinal hernia repair in 0- to 6-year-old children.

BACKGROUND/PURPOSE: This study provides epidemiologic data on the incidence of inguinal hernia repair in preschool children using the Taiwan National Health Insurance Research Database. We believe that the data on hernia repair in said database provide a close approximation of the true incidence of inguinal hernia in young children. METHOD: A cohort of 1,073,891 deidentified individuals was randomly selected from an insured population of 23 million. Subjects born during the period 1997-2004 were followed from birth to 6 years. The chi-square test and logistic regression modeling were used for statistical analyses. RESULT: A total of 92,308 individuals were born during the study period. Of these individuals, 3881 underwent hernia repairs. The cumulative incidence of hernia repair in children aged 0 to 6 years was 4.20%/7 years. The boy/girl ratio was 4.27:1 and the unilateral/bilateral ratio was 3.77:1. The incidence of hernia repair among boys was highest during the first year of life, but then decreased with age. In contrast, the incidence among girls remained stable during the first 6 years of life. Boys younger than 1 year had more bilateral repairs than boys in other age groups (p<0.0001) and girls had significantly more bilateral repairs than boys (p<0.0001). Subjects with a history of preterm birth also had a higher incidence of hernia repair than subjects who were born at full term (odds ratio=2.34, p<0.0001). CONCLUSION: Yearly incidence of hernia repair was obtained from a nationwide database. Some of the observations have not been reported elsewhere.

Development of a semantic-based search system for immunization knowledge.

This study developed and implemented a children's immunization management system with English and Traditional Chinese immunization ontology for semantic-based search of immunization knowledge. Parents and guardians are able to search vaccination-related information effectively. Jena Java Application Programming Interface (API) was used to search for synonyms and associated classes in this domain and then use them for searching by Google Search API. The searching results do not only contain suggested web links but also include a basic introduction to vaccine and related preventable diseases. Compared with the Google keyword-based search, over half of the 31 trial users prefer using semantic-based search of this system. Although the search runtime on this system is not as fast as well-known search engines such as Google or Yahoo, it can accurately focus on searching for child vaccination information to provide search results that better conform to the needs of users. Furthermore, the system is also one of the few health knowledge platforms that support Traditional Chinese semantic-based search.

Misleading health-related information promoted through video-based social media: anorexia on YouTube.

INTRODUCTION: The amount of information being uploaded onto social video platforms, such as YouTube, Vimeo, and Veoh, continues to spiral, making it increasingly difficult to discern reliable health information from misleading content. There are thousands of YouTube videos promoting misleading information about anorexia (eg, anorexia as a healthy lifestyle). OBJECTIVE: The aim of this study was to investigate anorexia-related misinformation disseminated through YouTube videos. METHODS: We retrieved YouTube videos related to anorexia using the keywords anorexia, anorexia nervosa, proana, and thinspo on October 10, 2011.Three doctors reviewed 140 videos with approximately 11 hours of video content, classifying them as informative, pro-anorexia, or others. By informative we mean content describing the health consequences of anorexia and advice on how to recover from it; by pro-anorexia we mean videos promoting anorexia as a fashion, a source of beauty, and that share tips and methods for becoming and remaining anorexic. The 40 most-viewed videos (20 informative and 20 pro-anorexia videos) were assessed to gauge viewer behavior. RESULTS: The interrater agreement of classification was moderate (Fleiss' kappa=0.5), with 29.3% (n=41) being rated as pro-anorexia, 55.7% (n=78) as informative, and 15.0% (n=21) as others. Pro-anorexia videos were favored 3 times more than informative videos (odds ratio [OR] 3.3, 95% CI 3.3-3.4, P<.001). CONCLUSIONS: Pro-anorexia information was identified in 29.3% of anorexia-related videos. Pro-anorexia videos are less common than informative videos; however, in proportional terms, pro-anorexia content is more highly favored and rated by its viewers. Efforts should focus on raising awareness, particularly among teenagers, about the trustworthiness of online information about beauty and healthy lifestyles. Health authorities producing videos to combat anorexia should consider involving celebrities and models to reach a wider audience. More research is needed to study the characteristics of pro-anorexia videos in order to develop algorithms that will automatically detect and filter those videos before they become popular.

Design of a Personal Health Record and Health Knowledge Sharing System using IHE-XDS and OWL.

Personal Health Record systems (PHRs) provide opportunities for patients to access their own PHR. However, PHRs are teeming with medical terminologies, such as disease and symptom names, etc. Patients need readily understandable and useful health knowledge in addition to their records in order to enhance their self-care ability. This study describes a Personal Health Record and Health Knowledge Sharing System (PHR&HKS) whereby users not only can maintain and import their PHR, but also can collate useful health Web resources that are related to their personal diseases. Furthermore, they can share the collated Web resources with any user with the same diseases and vice versa. To fulfill these objectives, IHE Cross-Enterprise Document Sharing (XDS) architecture was adopted to share and integrate the PHR. A registry ontology, consisting of part of the XDS document metadata attributes, the ICD-9-CM code, and part of the Dublin Core Metadata Element Set (DCMES), was created to enhance the health knowledge collating and sharing functions. The system was then tested and evaluated by 30 users. Among these individuals, 24 (81 %) held positive views on the ease of use and usefulness of the system while the remainder, who held either neutral (14 %) or negative (5 %) attitudes, were identified as individuals who were somewhat unwilling to maintain any PHR or share any information with others.

Telemedicine utilization to support the management of the burns treatment involving patient pathways in both developed and developing countries: a case study.

This case study reports on the utilization of telemedicine to support the management of the burns treatment in the islands of Sao Tome and Principe by Taipei Medical University-affiliated hospital in Taiwan. The authors share experiences about usage of telemedicine to support treatment of the burn victims in a low-income country that receive reconstructive surgery in a developed country. Throughout the entire care process, telemedicine has been used not only to provide an expert advice from distance but also to help establish and maintain the doctor-patient relationship, to keep patients in contact with their families, and to help educate and consult the medical personal physically present in Sao Tome and Principe. This case study presents the details of how this process has been conducted to date, on what were learned from this process, and on issues that should be considered to improve this process in the future. The authors plan to create instructional videos and post them on YouTube to aid clinical workers providing similar treatment during the acute care and rehabilitation process and also to support eLearning in many situations where it otherwise is not possible to use videoconferencing to establish real-time contact between doctors at the local site and remote specialists.

Study on the potential for delay tolerant networks by health workers in low resource settings.

BACKGROUND: Medical Informatics Systems (MIS) have been suggested as having great potential to improve health care delivery in low resource settings. One of the major barriers for adopting MIS in this context is a lack of adequate network/communication infrastructure. Delay Tolerant Networking (DTN) is an approach for establishing network connectivity in situations where it is possible to support physical transport of the digital information. To date most DTN research has been technically oriented, and very few services have been implemented to support healthcare systems using the technology. It is thus unclear about the potential that DTN may have for supporting MIS systems in low resource settings. The goals of the paper are twofold, first, to gain an initial estimate of interest in different services that can be supported by DTN. Second, to find out the necessary frequency associated with each service for supporting health work in low resource settings. METHOD: Fifty questionnaires were distributed to attendants at the International Conference on Global Health that had acknowledged having health work experience in a poor connectivity context. The respondents were using a 5-point Likert scale regarding if 9 different potential DTN services "would be useful". They also were asked how often data delivery would be necessary for these services to be useful. The Chi square was calculated to measure acceptance. RESULTS: 37 responses were received, aggregating the response rate of 74%. The respondents represented having work experience from 8 months to 15 years from 35 resource poor countries. The Chi square test showed very high statistical significance for "strongly agree and agree" for the potential usefulness of the proposed DTN services, with a p-value less than 0.001. The frequency of data delivery that would be necessary for services to be useful varied considerably. CONCLUSION: This study provides evidence of potential for DTN to support useful services that support health work in low resource settings, and that services like access to email, notification of lab results, backup of EHR and teleconsultation are seem to be most important services that can be supported by DTN. The necessary frequency of data delivery for each service, will be highly dependent on context. In a low resource setting with limited mobility, the physical transport of digital data at a frequency of less than once per week should still be sufficient for useful services like notification of lab results and ordering of medical supplies. Research comparing different methods for delivery of DTN data should thus be useful. Further research and collaboration between MIS and Computer Science research communities is recommended in order to help develop DTN services that can be evaluated. Efforts to enhance awareness among stakeholders about how DTN can be used to support health services should be worthwhile.

Magnetic resonance imaging for lung cancer detection: experience in a population of more than 10,000 healthy individuals.

BACKGROUND: Recent refinements of lung MRI techniques have reduced the examination time and improved diagnostic sensitivity and specificity. We conducted a study to assess the feasibility of MRI for the detection of primary lung cancer in asymptomatic individuals. METHODS: A retrospective chart review was performed on images of lung parenchyma, which were extracted from whole-body MRI examinations between October 2000 and December 2007. 11,766 consecutive healthy individuals (mean age, 50.4 years; 56.8% male) were scanned using one of two 1.5-T scanners (Sonata and Sonata Maestro, Siemens Medical Solutions, Erlangen, Germany). The standard protocol included a quick whole-lung survey with T2-weighted 2-dimensional half Fourier acquisition single shot turbo spin echo (HASTE) and 3-dimensional volumetric interpolated breath-hold examination (VIBE). Total examination time was less than 10 minutes, and scanning time was only 5 minutes. Prompt referrals and follow-ups were arranged in cases of suspicious lung nodules. RESULTS: A total of 559 individuals (4.8%) had suspicious lung nodules. A total of 49 primary lung cancers were diagnosed in 46 individuals: 41 prevalence cancers and 8 incidence cancers. The overall detection rate of primary lung cancers was 0.4%. For smokers aged 51 to 70 years, the detection rate was 1.4%. TNM stage I disease accounted for 37 (75.5%). The mean size of detected lung cancers was 1.98 cm (median, 1.5 cm; range, 0.5-8.2 cm). The most histological types were adenocarcinoma in 38 (77.6%). CONCLUSION: Rapid zero-dose MRI can be used for lung cancer detection in a healthy population.