Data mining to retrieve smoking status from electronic health records in general practice.

Aims: Optimize and assess the performance of an existing data mining algorithm for smoking status from hospital electronic health records (EHRs) in general practice EHRs. Methods and results: We optimized an existing algorithm in a training set containing all clinical notes from 498 individuals (75 712 contact moments) from the Julius General Practitioners' Network (JGPN). Each moment was classified as either 'current smoker', 'former smoker', 'never smoker', or 'no information'. As a reference, we manually reviewed EHRs. Algorithm performance was assessed in an independent test set (n = 494, 78 129 moments) using precision, recall, and F1-score. Test set algorithm performance for 'current smoker' was precision 79.7%, recall 78.3%, and F1-score 0.79. For former smoker, it was precision 73.8%, recall 64.0%, and F1-score 0.69. For never smoker, it was precision 92.0%, recall 74.9%, and F1-score 0.83. On a patient level, performance for ever smoker (current and former smoker combined) was precision 87.9%, recall 94.7%, and F1-score 0.91. For never smoker, it was 98.0, 82.0, and 0.89%, respectively. We found a more narrative writing style in general practice than in hospital EHRs. Conclusion: Data mining can successfully retrieve smoking status information from general practice clinical notes with a good performance for classifying ever and never smokers. Differences between general practice and hospital EHRs call for optimization of data mining algorithms when applied beyond a primary development setting.

Data mining information from electronic health records produced high yield and accuracy for current smoking status.

OBJECTIVES: Researchers are increasingly using routine clinical data for care evaluations and feedback to patients and clinicians. The quality of these evaluations depends on the quality and completeness of the input data. STUDY DESIGN AND SETTING: We assessed the performance of an electronic health record (EHR)-based data mining algorithm, using the example of the smoking status in a cardiovascular population. As a reference standard, we used the questionnaire from the Utrecht Cardiovascular Cohort (UCC). To assess diagnostic accuracy, we calculated sensitivity, specificity, negative predictive value (NPV), and positive predictive value (PPV). RESULTS: We analyzed 1,661 patients included in the UCC to January 18, 2019. Of those, 14% (n = 238) had missing information on smoking status in the UCC questionnaire. Data mining provided information on smoking status in 99% of the 1,661 participants. Diagnostic accuracy for current smoking was sensitivity 88%, specificity 92%, NPV 98%, and PPV 63%. From false positives, 85% reported they had quit smoking at the time of the UCC. CONCLUSION: Data mining showed great potential in retrieving information on smoking (a near complete yield). Its diagnostic performance is good for negative smoking statuses. The implications of misclassification with data mining are dependent on the application of the data.

eHealth as the Next-Generation Perinatal Care: An Overview of the Literature.

BACKGROUND: Unrestricted by time and place, electronic health (eHealth) provides solutions for patient empowerment and value-based health care. Women in the reproductive age are particularly frequent users of internet, social media, and smartphone apps. Therefore, the pregnant patient seems to be a prime candidate for eHealth-supported health care with telemedicine for fetal and maternal conditions. OBJECTIVE: This study aims to review the current literature on eHealth developments in pregnancy to assess this new generation of perinatal care. METHODS: We conducted a systematic literature search of studies on eHealth technology in perinatal care in PubMed and EMBASE in June 2017. Studies reporting the use of eHealth during prenatal, perinatal, and postnatal care were included. Given the heterogeneity in study methods, used technologies, and outcome measurements, results were analyzed and presented in a narrative overview of the literature. RESULTS: The literature search provided 71 studies of interest. These studies were categorized in 6 domains: information and eHealth use, lifestyle (gestational weight gain, exercise, and smoking cessation), gestational diabetes, mental health, low- and middle-income countries, and telemonitoring and teleconsulting. Most studies in gestational diabetes and mental health show that eHealth applications are good alternatives to standard practice. Examples are interactive blood glucose management with remote care using smartphones, telephone screening for postnatal depression, and Web-based cognitive behavioral therapy. Apps and exercise programs show a direction toward less gestational weight gain, increase in step count, and increase in smoking abstinence. Multiple studies describe novel systems to enable home fetal monitoring with cardiotocography and uterine activity. However, only few studies assess outcomes in terms of fetal monitoring safety and efficacy in high-risk pregnancy. Patients and clinicians report good overall satisfaction with new strategies that enable the shift from hospital-centered to patient-centered care. CONCLUSIONS: This review showed that eHealth interventions have a very broad, multilevel field of application focused on perinatal care in all its aspects. Most of the reviewed 71 articles were published after 2013, suggesting this novel type of care is an important topic of clinical and scientific relevance. Despite the promising preliminary results as presented, we accentuate the need for evidence for health outcomes, patient satisfaction, and the impact on costs of the possibilities of eHealth interventions in perinatal care. In general, the combination of increased patient empowerment and home pregnancy care could lead to more satisfaction and efficiency. Despite the challenges of privacy, liability, and costs, eHealth is very likely to disperse globally in the next decade, and it has the potential to deliver a revolution in perinatal care.