Background Evaluating patients' experiences is essential when incorporating the patients' perspective in improving healthcare. Experiences are mainly collected using closed-ended questions,... Show moreBackground Evaluating patients' experiences is essential when incorporating the patients' perspective in improving healthcare. Experiences are mainly collected using closed-ended questions, although the value of open-ended questions is widely recognized. Natural language processing (NLP) can automate the analysis of open-ended questions for an efficient approach to patient-centeredness. Methods We developed the Artificial Intelligence Patient-Reported Experience Measures (AI-PREM) tool, consisting of a new, open-ended questionnaire, an NLP pipeline to analyze the answers using sentiment analysis and topic modeling, and a visualization to guide physicians through the results. The questionnaire and NLP pipeline were iteratively developed and validated in a clinical context. Results The final AI-PREM consisted of five open-ended questions about the provided information, personal approach, collaboration between healthcare professionals, organization of care, and other experiences. The AI-PREM was sent to 867 vestibular schwannoma patients, 534 of which responded. The sentiment analysis model attained an F1 score of 0.97 for positive texts and 0.63 for negative texts. There was a 90% overlap between automatically and manually extracted topics. The visualization was hierarchically structured into three stages: the sentiment per question, the topics per sentiment and question, and the original patient responses per topic. Conclusions The AI-PREM tool is a comprehensive method that combines a validated, open-ended questionnaire with a well-performing NLP pipeline and visualization. Thematically organizing and quantifying patient feedback reduces the time invested by healthcare professionals to evaluate and prioritize patient experiences without being confined to the limited answer options of closed-ended questions. Show less
Heuvel, J.F.M. van den; Hogeveen, M.; Holzik, M.L.; Heijst, A.F.J. van; Bekker, M.N.; Geurtzen, R. 2022
Background In case of extreme premature delivery at 24 weeks of gestation, both early intensive care and palliative comfort care for the neonate are considered treatment options. Prenatal... Show moreBackground In case of extreme premature delivery at 24 weeks of gestation, both early intensive care and palliative comfort care for the neonate are considered treatment options. Prenatal counseling, preferably using shared decision making, is needed to agree on the treatment option in case labor progresses. This article described the development of a digital decision aid (DA) to support pregnant women, partners and clinicians in prenatal counseling for imminent extreme premature labor. Methods This DA is developed following the International Patient Decision Aid Standards. The Dutch treatment guideline and the Dutch recommendations for prenatal counseling in extreme prematurity were used as basis. Development of the first prototype was done by expert clinicians and patients, further improvements were done after alpha testing with involved clinicians, patients and other experts (n = 12), and beta testing with non-involved clinicians and patients (n = 15). Results The final version includes information, probabilities and figures depending on users' preferences. Furthermore, it elicits patient values and provides guidance to aid parents and professionals in making a decision for either early intensive care or palliative comfort care in threatening extreme premature delivery. Conclusion A decision aid was developed to support prenatal counseling regarding the decision on early intensive care versus palliative comfort care in case of extreme premature delivery at 24 weeks gestation. It was well accepted by parents and healthcare professionals. Our multimedia, digital DA is openly available online to support prenatal counseling and personalized, shared decision-making in imminent extreme premature labor. Show less
Egmond, M.B. van; Spini, G.; Galien, O. van der; IJpma, A.; Veugen, T.; Kraaij, W.; ... ; Kooij-Janic, M. 2021
Background Recent developments in machine learning have shown its potential impact for clinical use such as risk prediction, prognosis, and treatment selection. However, relevant data are often... Show moreBackground Recent developments in machine learning have shown its potential impact for clinical use such as risk prediction, prognosis, and treatment selection. However, relevant data are often scattered across different stakeholders and their use is regulated, e.g. by GDPR or HIPAA. As a concrete use-case, hospital Erasmus MC and health insurance company Achmea have data on individuals in the city of Rotterdam, which would in theory enable them to train a regression model in order to identify high-impact lifestyle factors for heart failure. However, privacy and confidentiality concerns make it unfeasible to exchange these data. Methods This article describes a solution where vertically-partitioned synthetic data of Achmea and of Erasmus MC are combined using Secure Multi-Party Computation. First, a secure inner join protocol takes place to securely determine the identifiers of the patients that are represented in both datasets. Then, a secure Lasso Regression model is trained on the securely combined data. The involved parties thus obtain the prediction model but no further information on the input data of the other parties. Results We implement our secure solution and describe its performance and scalability: we can train a prediction model on two datasets with 5000 records each and a total of 30 features in less than one hour, with a minimal difference from the results of standard (non-secure) methods. Conclusions This article shows that it is possible to combine datasets and train a Lasso regression model on this combination in a secure way. Such a solution thus further expands the potential of privacy-preserving data analysis in the medical domain. Show less
Background The use of statins for primary prevention of cardiovascular diseases is associated with different benefit and harm outcomes. The aime of this study is how important these outcomes are... Show moreBackground The use of statins for primary prevention of cardiovascular diseases is associated with different benefit and harm outcomes. The aime of this study is how important these outcomes are for people and what people's preferences are. Methods We conducted a preference-eliciting survey incorporating a best-worst scaling (BWS) instrument in Iran from June to November 2019. The relative importance of 13 statins-related outcomes was assessed on a sample of 1085 participants, including 913 general population (486 women) and 172 healthcare providers from the population covered by urban and rural primary health care centers. The participants made trade-off decisions and selected the most and least worrisome outcomes concurrently from 13 choice sets; each contains four outcomes generated using the balanced incomplete block design. Results According to the mean (SD) BWS scores, which can be (+ 4) in maximum and (- 4) in minimum, in the general population, the most worrisome outcomes were severe stroke (3.37 (0.8)), severe myocardial infarction (2.71(0.7)), and cancer (2.69 (1.33)). While myopathy (- 3. 03 (1.03)), nausea/headache (- 2.69 (0.94)), and treatment discontinuation due to side effects (- 2.24 (1.14)) were the least worrisome outcomes. Preferences were similar between rural and urban areas and among health care providers and the general population with overlapping uncertainty intervals. Conclusion The rank of health outcomes may be similar in various socio-cultural contexts. The preferences for benefits and harms of statin therapy are essential to assess benefit-harm balance when recommending statins for primary prevention of cardiovascular diseases. Show less
Berger, F.A.; Sijs, H. van der; Becker, M.L.; Gelder, T. van; Bemt, P.M.L.A. van den 2020
Background The exact risk of developing QTc-prolongation when using a combination of QTc-prolonging drugs is still unknown, making it difficult to interpret these QT drug-drug interactions (QT-DDIs... Show moreBackground The exact risk of developing QTc-prolongation when using a combination of QTc-prolonging drugs is still unknown, making it difficult to interpret these QT drug-drug interactions (QT-DDIs). A tool to identify high-risk patients is needed to support healthcare providers in handling automatically generated alerts in clinical practice. The main aim of this study was to develop and validate a tool to assess the risk of QT-DDIs in clinical practice. Methods A model was developed based on risk factors associated with QTc-prolongation determined in a prospective study on QT-DDIs in a university medical center inthe Netherlands. The main outcome measure was QTc-prolongation defined as a QTc interval > 450 ms for males and > 470 ms for females. Risk points were assigned to risk factors based on their odds ratios. Additional risk factors were added based on a literature review. The ability of the model to predict QTc-prolongation was validated in an independent dataset obtained from a general teaching hospital against QTc-prolongation as measured by an ECG as the gold standard. Sensitivities, specificities, false omission rates, accuracy and Youden's index were calculated. Results The model included age, gender, cardiac comorbidities, hypertension, diabetes mellitus, renal function, potassium levels, loop diuretics, and QTc-prolonging drugs as risk factors. Application of the model to the independent dataset resulted in an area under the ROC-curve of 0.54 (95% CI 0.51-0.56) when QTc-prolongation was defined as > 450/470 ms, and 0.59 (0.54-0.63) when QTc-prolongation was defined as > 500 ms. A cut-off value of 6 led to a sensitivity of 76.6 and 83.9% and a specificity of 28.5 and 27.5% respectively. Conclusions A clinical decision support tool with fair performance characteristics was developed. Optimization of this tool may aid in assessing the risk associated with QT-DDIs. Show less
Cammel, S.A.; Vos, M.S. de; Soest, D. van; Hettne, K.M.; Boer, F.; Steyerberg, E.W.; Boosman, H. 2020
Background Patient experience surveys often include free-text responses. Analysis of these responses is time-consuming and often underutilized. This study examined whether Natural Language... Show moreBackground Patient experience surveys often include free-text responses. Analysis of these responses is time-consuming and often underutilized. This study examined whether Natural Language Processing (NLP) techniques could provide a data-driven, hospital-independent solution to indicate points for quality improvement. Methods This retrospective study used routinely collected patient experience data from two hospitals. A data-driven NLP approach was used. Free-text responses were categorized into topics, subtopics (i.e. n-grams) and labelled with a sentiment score. The indicator 'impact', combining sentiment and frequency, was calculated to reveal topics to improve, monitor or celebrate. The topic modelling architecture was tested on data from a second hospital to examine whether the architecture is transferable to another hospital. Results A total of 38,664 survey responses from the first hospital resulted in 127 topics and 294 n-grams. The indicator 'impact' revealed n-grams to celebrate (15.3%), improve (8.8%), and monitor (16.7%). For hospital 2, a similar percentage of free-text responses could be labelled with a topic and n-grams. Between-hospitals, most topics (69.7%) were similar, but 32.2% of topics for hospital 1 and 29.0% of topics for hospital 2 were unique. Conclusions In both hospitals, NLP techniques could be used to categorize patient experience free-text responses into topics, sentiment labels and to define priorities for improvement. The model's architecture was shown to be hospital-specific as it was able to discover new topics for the second hospital. These methods should be considered for future patient experience analyses to make better use of this valuable source of information. Show less
Claassen, A.A.O.M.; Schers, H.J.; Busch, V.J.J.F.; Heesterbeek, P.J.C.; Hoogen, F.H.J. van den; Vlieland, T.P.M.V.; Ende, C.H.M. van den 2020
Background To evaluate the effect of a stand-alone mobile and web-based educational intervention (eHealth tool) compared to usual preparation of a first orthopedic consultation of patients with hip... Show moreBackground To evaluate the effect of a stand-alone mobile and web-based educational intervention (eHealth tool) compared to usual preparation of a first orthopedic consultation of patients with hip or knee osteoarthritis (OA) on patients' satisfaction. Methods A two-armed randomized controlled trial involving 286 patients with (suspicion of) hip or knee OA, randomly allocated to either receiving an educational eHealth tool to prepare their upcoming consultation (n = 144) or usual care (n = 142). Satisfaction with the consultation on three subscales (range 1-4) of the Consumer Quality Index (CQI - primary outcome) and knowledge (assessed using 22 statements on OA, range 0-22), treatment beliefs (assessed by the Treatment beliefs in OsteoArthritis questionnaire, range 1-5), assessment of patient's involvement in consultation by the surgeon (assessed on a 5-point Likert scale) and patient satisfaction with the outcome of the consultation (numeric rating scale), were assessed. Results No differences between groups were observed on the 3 subscales of the CQI (group difference (95% CI): communication 0.009 (- 0.10, 0.12), conduct - 0.02 (- 0.12, 0.07) and information provision 0.02 (- 0.18, 0.21)). Between group differences (95% CI) were in favor of the intervention group for knowledge (1.4 (0.6, 2.2)), negative beliefs regarding physical activities (- 0.19 (- 0.37, - 0.002) and pain medication (- 0.30 (- 0.49, - 0.01)). We found no differences on other secondary outcomes. Conclusions An educational eHealth tool to prepare a first orthopedic consultation for hip or knee OA does not result in higher patient satisfaction with the consultation, but it does influence cognitions about osteoarthritis. Show less
Bakker, T.; Klopotowska, J.E.; Eslami, S.; Lange, D.W. de; Marum, R. van; Sijs, H. van der; ... ; Abu-Hanna, A. 2019
BACKGROUND\nWe aimed to evaluate the effect of a decision aid (DA) with information only compared to a DA with values clarification exercise (VCE), and to study the role of personality and... Show moreBACKGROUND\nWe aimed to evaluate the effect of a decision aid (DA) with information only compared to a DA with values clarification exercise (VCE), and to study the role of personality and information seeking style in DA-use, decisional conflict (DC) and knowledge.\nMETHODS\nTwo scenario-based experiments were conducted with two different groups of healthy female participants. Dependent measures were: DC, knowledge, and DA-use (time spent, pages viewed, VCE used). Respondents were randomized between a DA with information only (VCE-) and a DA with information plus a VCE(VCE+) (experiment 1), or between information only (VCE-), information plus VCE without referral to VCE(VCE+), and information plus a VCE with specific referral to the VCE, requesting participants to use the VCE(VCE++) (experiment 2). In experiment 2 we additionally measured personality (neuroticism/conscientiousness) and information seeking style (monitoring/blunting).\nRESULTS\nExperiment 1. There were no differences in DC, knowledge or DA-use between VCE- (n=70) and VCE+ (n=70). Both DAs lead to a mean gain in knowledge from 39% at baseline to 73% after viewing the DA. Within VCE+, VCE-users (n=32, 46%) reported less DC compared to non-users. Since there was no difference in DC between VCE- and VCE+, this is likely an effect of VCE-use in a self-selected group, and not of the VCE per se. Experiment 2. There were no differences in DC or knowledge between VCE- (n=65), VCE+ (n=66), VCE++ (n=66). In all groups, knowledge increased on average from 42% at baseline to 72% after viewing the DA. Blunters viewed fewer DA-pages (R=0.38, p<.001). More neurotic women were less certain (R=0.18, p<.01) and felt less supported in decision making (R=0.15, p<.05); conscientious women felt more certain (R=-0.15, p<.05) and had more knowledge after viewing the DA (R=0.15, p<.05).\nCONCLUSIONS\nBoth DAs lead to increased knowledge in healthy populations making hypothetical decisions, and use of the VCE did not improve knowledge or DC. Personality characteristics were associated to some extent with DA-use, information seeking styles with aspects of DC. More research is needed to make clear recommendations regarding the need for tailoring of information provision to personality characteristics, and to assess the effect of VCE use in actual patients. Show less