Dataset Information

A Data-Driven Iterative Approach for Semi-automatically Assessing the Correctness of Medication Value Sets: A Proof of Concept Based on Opioids.

ABSTRACT:

Background

Value sets are lists of terms (e.g., opioid medication names) and their corresponding codes from standard clinical vocabularies (e.g., RxNorm) created with the intent of supporting health information exchange and research. Value sets are manually-created and often exhibit errors.

Objectives

The aim of the study is to develop a semi-automatic, data-centric natural language processing (NLP) method to assess medication-related value set correctness and evaluate it on a set of opioid medication value sets.

Methods

We developed an NLP algorithm that utilizes value sets containing mostly true positives and true negatives to learn lexical patterns associated with the true positives, and then employs these patterns to identify potential errors in unseen value sets. We evaluated the algorithm on a set of opioid medication value sets, using the recall, precision and F₁-score metrics. We applied the trained model to assess the correctness of unseen opioid value sets based on recall. To replicate the application of the algorithm in real-world settings, a domain expert manually conducted error analysis to identify potential system and value set errors.

Results

Thirty-eight value sets were retrieved from the Value Set Authority Center, and six (two opioid, four non-opioid) were used to develop and evaluate the system. Average precision, recall, and F₁-score were 0.932, 0.904, and 0.909, respectively on uncorrected value sets; and 0.958, 0.953, and 0.953, respectively after manual correction of the same value sets. On 20 unseen opioid value sets, the algorithm obtained average recall of 0.89. Error analyses revealed that the main sources of system misclassifications were differences in how opioids were coded in the value sets-while the training value sets had generic names mostly, some of the unseen value sets had new trade names and ingredients.

Conclusion

The proposed approach is data-centric, reusable, customizable, and not resource intensive. It may help domain experts to easily validate value sets.

SUBMITTER: Li L

PROVIDER: S-EPMC8716187 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:BackgroundPostoperative recovery after abdominal surgery is measured mostly based on subjective or self-reported data. In this article we aim to evaluate whether recovery of daily physical activity levels can be measured postoperatively with the use of an accelerometer.MethodsIn this multicenter, observational pilot study, 30 patients undergoing laparoscopic abdominal surgery (hysterectomy, adnexal surgery, cholecystectomy and hernia inguinal surgery) were included. Patients were instructed to wear an Actigraph wGT3X-BT accelerometer during one week before surgery (baseline) and during the first, third and fifth week after surgery. Wear time, steps taken and physical activity intensity levels (sedentary, light, moderate and vigorous) were measured. Patients were blinded for the accelerometer outcomes. Additionally, an activity diary comprising patients' self-reported time of being recovered and a list of 18 activities, in which the dates of resumption of these 18 activities were recorded after surgery, was completed by the patient.ResultsFive patients were excluded from analyses because of technical problems with the accelerometer (n = 1) and protocol non-adherence (n = 4). Light, moderate, vigorous, combined moderate and vigorous intensity physical activity (MVPA), and step counts showed a clear recovery curve after surgery. Patients who underwent minor surgery reached their baseline step count and MVPA three weeks after surgery. Patients who underwent intermediate surgery had not yet reached their baseline step count during the last measuring week (five weeks after surgery). The results of the activity diaries showed a fair agreement with the accelerometer results (Cohens Kappa range: 0.273-0.391). Wearing the accelerometer was well tolerated and not regarded as being burdensome by the patients.ConclusionsThe accelerometer appeared to be a feasible way to measure recovery of postoperative physical activity levels in this study and was well tolerated by the patients. The agreement with self-reported physical recovery times was fair.

Project description:Infant sociability is generally conceived in terms of dyadic capacities and behaviors. Recently, quantitative evidence has been published to support arguments that infants achieve a criterion for groupness: the capacity to interact simultaneously with two others. Such studies equate this capacity with alternating dyadic acts to the two other members of an interacting trio. Here we propose a stricter threefold criterion for infant groupness, of which the crux is whether the social behavior of an infant at time B is shown to be influenced by what two or more group-members were previously doing at time A. We test the viability of this conceptualization: (a) through its justification of the novel laboratory procedure of studying infant sociability in infant-peer quartets (rather than trios); and, (b) in an analysis of a pilot study of gaze-behavior recorded in 5-min interactions among two quartets of infants aged 6-9 months. We call this a 'proof of concept' because our aim is to show that infants are capable of groupness, when groupness is conceptualized in a supra-dyadic way-not that all infants will manifest it, nor that all conditions will produce it, nor that it is commonplace in infants' everyday lives. We found that both quartets did achieve the minimum criterion of groupness that we propose: mutual gaze predicting coordinated gaze (where two babies, A and B, are looking at each other, and B is then looked at by C, and sometimes D) more strongly than the reverse. There was a significant absence of 'parallel mutual gaze,' where the four babies pair off. We conclude that, under specific conditions, preverbal infants can manifest supra-dyadic groupness. Infants' capacities to exhibit groupness by 9 months of age, and the paucity of parallel mutual gaze in our data, run counter to the assumption that infant sociability, when in groups, is always generated by a dyadic program. Our conceptualization and demonstration of groupness in 8-month-olds thus opens a host of empirical, theoretical, and practical questions about the sociability and care of young babies.

Project description:BackgroundMobile phone based programs for kidney transplant recipients are promising tools for improving long-term graft outcomes and better managing comorbidities (eg, hypertension, diabetes). These tools provide an easy to use self-management framework allowing optimal medication adherence that is guided by the patients' physiological data. This technology is also relatively inexpensive, has an intuitive interface, and provides the capability for real-time personalized feedback to help motivate patient self-efficacy. Automated summary reports of patients' adherence and blood pressure can easily be uploaded to providers' networks helping reduce clinical inertia by reducing regimen alteration time.ObjectiveThe aim of this study was to assess the feasibility, acceptability, and preliminary outcomes of a prototype mobile health (mHealth) medication and blood pressure (BP) self-management system for kidney transplant patients with uncontrolled hypertension.MethodsA smartphone enabled medication adherence and BP self-management system was developed using a patient and provider centered design. The development framework utilized self-determination theory with iterative stages that were guided and refined based on patient/provider feedback. A 3-month proof-of-concept randomized controlled trial was conducted in 20 hypertensive kidney transplant patients identified as non-adherent to their current medication regimen based on a month long screening using an electronic medication tray. Participants randomized to the mHealth intervention had the reminder functions of their electronic medication tray enabled and received a bluetooth capable BP monitor and a smartphone that received and transmitted encrypted physiological data and delivered reminders to measure BP using text messaging. Controls received standard of care and their adherence continued to be monitored with the medication tray reminders turned off. Providers received weekly summary reports of patient medication adherence and BP readings.ResultsParticipation and retention rates were 41/55 (75%) and 31/34 (91%), respectively. The prototype system appears to be safe, highly acceptable, and useful to patients and providers. Compared to the standard care control group (SC), the mHealth intervention group exhibited significant improvements in medication adherence and significant reductions in clinic-measured systolic blood pressures across the monthly evaluations. Physicians made more anti-hypertensive medication adjustments in the mHealth group versus the standard care group (7 adjustments in 5 patients versus 3 adjustments in 3 patients) during the 3-month trial based on the information provided in the weekly reports.ConclusionsThese data support the acceptability and feasibility of the prototype mHealth system. Further trials with larger sample sizes and additional biomarkers (eg, whole blood medication levels) are needed to examine efficacy and effectiveness of the system for improving medication adherence and blood pressure control after kidney transplantation over longer time periods.Trial registrationClinicaltrials.gov NCT01859273; http://clinicaltrials.gov/ct2/show/NCT01859273 (Archived by WebCite at http://www.webcitation.org/6IqfCa3A3).

Project description:Behavioral pharmacology paradigms have been used for early efficacy testing of novel compounds for alcohol use disorder (AUD). However, the degree to which early efficacy in the human laboratory predicts clinical efficacy remains unclear. To address this gap in the literature we employed a novel meta-analytic approach. We searched the literature for medications tested for AUD using both behavioral pharmacology (i.e., alcohol administration) and randomized clinical trials (RCTs). For behavioral pharmacology, we computed medication effects on alcohol-induced stimulation, sedation, and craving during the alcohol administration (k = 51 studies, 24 medications). For RCTs, we computed medication effects on any drinking and heavy drinking (k = 118 studies, 17 medications). We used medication as the unit of analysis and applied the Williamson-York bivariate weighted least squares estimation to preserve the errors in both the independent and dependent variables. Results, with correction for publication bias, revealed a significant and positive relationship between medication effects on alcohol-induced stimulation (β = 1.18 p < 0.05), sedation (β = 2.38, p < 0.05), and craving (β = 3.28, p < 0.001) in the laboratory, and drinking outcomes in RCTs, such that medications that reduced stimulation, sedation, and craving during the alcohol administration were associated with better clinical outcomes. A leave-one-out Monte Carlo analysis examined the predictive utility of these laboratory endpoints for each medication. The observed clinical effect size was within one standard deviation of the mean predicted effect size for all but three pharmacotherapies. This proof-of-concept study demonstrates that behavioral pharmacology endpoints of alcohol-induced stimulation, sedation, and craving track medication effects from the human laboratory to clinical trial outcomes. These results apply to alcohol administration phenotypes and may be especially useful to medications for which the mechanisms of action involve alterations in subjective responses to alcohol (e.g., antagonist medication). These methods and results can be applied to a host of clinical questions and can streamline the process of screening novel compounds for AUD. For instance, this approach can be used to quantify the predictive utility of cue-reactivity screening models and even preclinical models of medication development.