Predictors of medication adherence in elderly patients with chronic diseases using support vector machine models

The Morisky Method for Measuring Medication Adherence in Older Adults With Chronic Diseases: A Cross-Sectional Study

Background and Aims

In the elderly population, the prevalence of chronic diseases and the necessity for supportive and medication treatments are increasing, making medication adherence a crucial factor in enhancing their quality of life.

Methods

This study conducted a comprehensive descriptive-analytical survey on older adults aged over 65 with chronic diseases, who were visiting a specialized diabetes clinic in YAZD in 2023. The clinic provides treatment for elderly diabetic patients as well as those with chronic diseases resulting from or concurrent with diabetes, managed by internal medicine specialists and endocrinologists. The participants had been taking medication for more than 6 months and were suffering from chronic conditions such as asthma, hypertension, diabetes, chronic cardiovascular disease, liver cirrhosis, stroke, and vascular heart disease, with normal cognitive function. Medication adherence was assessed to determine the level of adherence. Data were analyzed using SPSS version 20.0, utilizing logistic regression.

Results

A total of 196 participants took part in the study. The average medication knowledge score was 14.7 ± 3.5, the average depression score was 8.1 ± 2.4, the average health literacy score was 7.5 ± 1.6, and the average self-efficacy score was 29.1 ± 5. Logistic regression analysis revealed that more than half of the participants (58.7%) lacked medication adherence. The analysis also indicated that the presence of a spouse had a significant effect on medication adherence (p-value = 0.038), along with health literacy (p-value = 0.002) and self-efficacy (p-value = 0.000), which had the most significant impact on medication adherence.

Conclusion

The findings suggest that self-efficacy, health literacy, and the presence of a spouse are crucial factors influencing medication adherence in older adults with chronic diseases. These factors can shape the beliefs, attitudes, and behaviors of older adults regarding medication adherence and affect their health outcomes and quality of life. Therefore, interventions aimed at improving medication adherence in this population should consider these factors and attention to the specific needs and preferences of older adults and their spouses or family members.

Determinants of medication adherence among elderly with high blood pressure living in deprived areas

INTRODUCTION

The current study was conducted to determine the impact of health literacy and factors related to adherence to drug treatment, using the model proposed by the World Health Organization, in older adults with hypertension residing in informal settlements in Hamadan.

METHODS

This cross-sectional study was conducted on 405 patients in Hamadan city, located in the western part of Iran. Data were collected using an interviewer-administered questionnaire that included the 5-dimensional model proposed by the World Health Organization, Health Literacy for Iranian Adults, and Morisky Medication Adherence Scale-8. A two-stage sampling procedure was used to select patients from 14 comprehensive health service centers and health bases. The data were analyzed using SPSS v.24.

RESULTS

The study found that medication adherence was suboptimal in 63% of the participants. Additionally, 87.5% of patients had inadequate or insufficient health literacy. Factors related to medication adherence included age (odds ratio (OR) = 1.07), annual income (OR = 0.17), duration of hypertension (OR = 7.33), health literacy (OR = 1.03), self-reported health status (P < 0.05), and regular medication use (P < 0.008).

CONCLUSION

The results of this study indicate that more than half of the older adults in the study had suboptimal medication adherence and insufficient health literacy. The study also found that various factors, such as socioeconomic status, disease and treatment-related factors, and patient-related factors, influence medication adherence among older adults.

Prediction of 30-day in-hospital mortality in older UGIB patients using a simplified risk score and comparison with AIMS65 score

BACKGROUND

Upper gastrointestinal bleeding (UGIB) in older patients is associated with substantial in-hospital morbidity and mortality. This study aimed to develop and validate a simplified risk score for predicting 30-day in-hospital mortality in this population.

METHODS

A retrospective analysis was conducted on data from 1899 UGIB patients aged ≥ 65 years admitted to a single medical center between January 2010 and December 2019. An additional cohort of 330 patients admitted from January 2020 to October 2021 was used for external validation. Variable selection was performed using five distinct methods, and models were generated using generalized linear models, random forest, support vector machine, and k-nearest neighbors approaches. The developed score, "ABCAP," incorporated Albumin < 30 g/L, Blood Urea Nitrogen (BUN) > 7.5 mmol/L, Cancer presence, Altered mental status, and Pulse rate > 100/min, each assigned a score of 1. Internal and external validation procedures compared the ABCAP score with the AIMS65 score.

RESULTS

In internal validation, the ABCAP score demonstrated robust predictive capability with an area under the curve (AUC) of 0.878 (95% CI: 0.824-0.932), which was significantly better than the AIMS65 score (AUC: 0.827, 95% CI: 0.751-0.904), as revealed by the DeLong test (p = 0.048). External validation of the ABCAP score resulted in an AUC of 0.799 (95% CI: 0.709-0.889), while the AIMS65 score yielded an AUC of 0.743 (95% CI: 0.647-0.838), with no significant difference between the two scores based on the DeLong test (p = 0.16). However, the ABCAP score at the 3-5 score level demonstrated superior performance in identifying high-risk patients compared to the AIMS65 score. This score exhibited consistent predictive accuracy across variceal and non-variceal UGIB subgroups.

CONCLUSIONS

The ABCAP score incorporates easily obtained clinical variables and demonstrates promising predictive ability for 30-day in-hospital mortality in older UGIB patients. It allows effective mortality risk stratification and showed slightly better performance than the AIMS65 score. Further cohort validation is required to confirm generalizability.

Effects of a medication adherence app among medically underserved adults with chronic illness: a randomized controlled trial

For individuals living with a chronic illness who require use of long-term medications, adherence is a vital aspect of successful symptom management and outcomes. This study investigated the effect of a smartphone app on adherence, self-efficacy, knowledge, and medication social support in a medically underserved adult population with various chronic illnesses. Participants were randomized to a group who used the app for one month or a control group provided with a printed medication list. Compared to the control group, participants receiving the intervention had significantly greater medication adherence (Cohen's d = -0.52, p = .014) and medication self-efficacy (Cohen's d = 0.43, p = .035). No significant effects were observed related to knowledge or social support. The findings suggest use of the app could positively impact chronic disease management in a medically underserved population in the United States.

Effect of a Medication Adherence Mobile Phone App on Medically Underserved Patients with Chronic Illness: Preliminary Efficacy Study

BACKGROUND

Medication adherence is vital in the treatment of patients with chronic illness who require long-term medication therapies to maintain optimal health. Medication adherence, a complex and widespread problem, has been difficult to solve. Additionally, lower-income, medically underserved communities have been found to have higher rates of inadequate adherence to oral medications. Even so, this population has been underrepresented in studies using mobile medication adherence app interventions. Federally qualified health centers provide care for medically underserved populations, defined as communities and populations where there is a demonstrable unmet need for health services. These centers have been reporting an increase in a more complex chronic disease population. Including medically underserved individuals in mobile health studies provides opportunities to support this disproportionately affected group, work toward reducing health disparities in access to health care, and understand barriers to mobile health uptake.

OBJECTIVE

The aim of this preliminary efficacy study was to evaluate the effects and feasibility of a commercially available medication adherence app, Medisafe, in a medically underserved adult population with various chronic illnesses seeking care in a federally qualified health center.

METHODS

Participants in this single-arm pre-post intervention preliminary efficacy study (N=10) completed a baseline survey, used the app for 2 weeks, and completed an end-of-study survey. The primary outcome measures were medication adherence and medication self-efficacy. Feedback on the use of the app was also gathered.

RESULTS

A statistically significant median increase of 8 points on the self-efficacy for adherence to medications scale was observed (P=.03, Cohen d=0.69). Though not significant, the adherence to refills and medications scale demonstrated a median change of 2.5 points in the direction of increased medication adherence (P=.21, Cohen d=0.41). Feedback about the app was positive.

CONCLUSIONS

Use of the Medisafe app is a viable option to improve medication self-efficacy and medication adherence in medically underserved patients in an outpatient setting with a variety of chronic illnesses.

Relationship between People's Interest in Medication Adherence, Health Literacy, and Self-Care: An Infodemiological Analysis in the Pre- and Post-COVID-19 Era

The prevalence of non-communicable diseases has risen sharply in recent years, particularly among older individuals who require complex drug regimens. Patients are increasingly required to manage their health through medication adherence and self-care, but about 50% of patients struggle to adhere to prescribed treatments. This study explored the relationship between interest in medication adherence, health literacy, and self-care and how it changed during the COVID-19 pandemic. We used Google Trends to measure relative search volumes (RSVs) for these three topics from 2012 to 2022. We found that interest in self-care increased the most over time, followed by health literacy and medication adherence. Direct correlations emerged between RSVs for medication adherence and health literacy (r = 0.674, p < 0.0001), medication adherence and self-care (r = 0.466, p < 0.0001), and health literacy and self-care (r = 0.545, p < 0.0001). After the COVID-19 pandemic outbreak, interest in self-care significantly increased, and Latin countries showed a greater interest in self-care than other geographical areas. This study suggests that people are increasingly interested in managing their health, especially in the context of the recent pandemic, and that infodemiology may provide interesting information about the attitudes of the population toward chronic disease management.

Self-reported health literacy and medication adherence in older adults: a systematic review

OBJECTIVES

To give an overview over the associations between self-reported health literacy and medication adherence in older adults.

DESIGN

A systematic literature review of quantitative studies published in English and German.

DATA SOURCES

MEDLINE via PubMed, CINAHL, Cochrane Library, Epistemonikos and LIVIVO were searched.

ELIGIBILITY CRITERIA

Included studies had to examine the associations between self-reported health literacy and medication adherence in the elderly (samples including ≥66% of ≥60 years old) and had to use a quantitative methodology and had to be written in English or German.

DATA EXTRACTION AND SYNTHESIS

All studies were screened for inclusion criteria by two independent reviewers. A narrative synthesis was applied to analyse all included studies thematically. Quality assessment was conducted using the NIH Quality Assessment Tool for Observational Cohort and Cross-Sectional Studies.

RESULTS

We found 2313 studies, of which nine publications from eight studies were included in this review. Five studies reported a majority of participants with limited health literacy, one study reported a majority of participants with adequate health literacy, and three publications from two studies only reported mean levels of health literacy. Eight publications from seven studies used self-reports to measure medication adherence, while one study used the medication possession ratio. Overall, six publications from five studies reported significantly positive associations between health literacy and medication adherence while two studies reported positive but non-significant associations between both constructs and one study reported mixed results.

CONCLUSION

In this review, associations between self-reported health literacy and medication adherence are rather consistent, indicating positive associations between both constructs in older adults. However, concepts and measures of health literacy and medication adherence applied in the included studies still show a noteworthy amount of heterogeneity (eg, different use of cutoffs). These results reveal the need for more differentiated research in this area.

PROSPERO REGISTRATION NUMBER

CRD42019141028.

Machine Learning to Predict Tamoxifen Nonadherence Among US Commercially Insured Patients With Metastatic Breast Cancer

PURPOSE

Adherence to tamoxifen citrate among women diagnosed with metastatic breast cancer can improve survival and minimize recurrence. This study aimed to use real-world data and machine learning (ML) methods to classify tamoxifen nonadherence.

METHODS

A cohort of women diagnosed with metastatic breast cancer from 2012 to 2017 were identified from IBM MarketScan Commercial Claims and Encounters and Medicare claims databases. Patients with < 80% proportion of days coverage in the year following treatment initiation were classified as nonadherent. Training and internal validation cohorts were randomly generated (4:1 ratio). Clinical procedures, comorbidity, treatment, and health care encounter features in the year before tamoxifen initiation were used to train logistic regression, boosted logistic regression, random forest, and feedforward neural network models and were internally validated on the basis of area under receiver operating characteristic curve. The most predictive ML approach was evaluated to assess feature importance.

RESULTS

A total of 3,022 patients were included with 40% classified as nonadherent. All models had moderate predictive accuracy. Logistic regression (area under receiver operating characteristic 0.64) was interpreted with 94% sensitivity (95% CI, 89 to 92) and 0.31 specificity (95% CI, 29 to 33). The model accurately classified adherence (negative predictive value 89%) but was nondiscriminate for nonadherence (positive predictive value 48%). Variable importance identified top predictive factors, including age ≥ 55 years and pretreatment procedures (lymphatic nuclear medicine, radiation oncology, and arterial surgery).

CONCLUSION

ML using baseline administrative data predicts tamoxifen nonadherence. Screening at treatment initiation may support personalized care, improve health outcomes, and minimize cost. Baseline claims may not be sufficient to discriminate adherence. Further validation with enriched longitudinal data may improve model performance.

Exploring Barriers to Medication Adherence Using COM-B Model of Behaviour Among Patients with Cardiovascular Diseases in Low- and Middle-Income Countries: A Qualitative Study

INTRODUCTION

In 2016, cardiovascular diseases (CVDs) led to 17.9 million deaths worldwide, representing 31% of all global deaths. CVDs are the leading cause of mortality worldwide and significant barriers to achieving the sustainable development goals. Modern medicines have been significant in improving health outcomes. However, non-adherence to medication is one of the reasons behind adverse health-related outcomes among patients suffering from atherosclerotic cardiovascular disease in low- and middle-income countries.

PATIENTS AND METHODS

This qualitative study was conducted at two tertiary care hospitals in India and Ghana. A total of 35 in-depth interviews were conducted with atherosclerosis cardiovascular disease (ASCVD) patients. The data were analysed thematically using the Capability Opportunity and Motivation (COM-B) framework.

FINDINGS

The findings were summarised under three important broad themes of the COM-B framework: capability, opportunity and behaviour. Under capability, comprehension of disease, medication schedule, and unplanned travel affected adherence among patients. Cost of medication, insurance and access were the critical factors under opportunity, which negatively influenced medication adherence. Mood, beliefs about treatment and outcome expectations under motivation led to non-adherence among patients. Apart from these factors, some important health system factors such as health care experience and trust in the facilities and reliance on alternative medication also affected adherence in both countries.

CONCLUSION

This study has highlighted that the health system factors have dominantly influenced adherence to medication in India and Ghana. In India, we found participants to be satisfied with their health care provided at the government hospitals. However, limited time for consultation, lack of well-stocked pharmacy and unclear prescription negatively influenced adherence among participants in India and Ghana. The study emphasises that the health system needs to be strengthened, and the patients' belief system needs to be explored to address the issue of medication adherence in LMICs.

Application of machine-learning techniques in classification of HIV medical care status for people living with HIV in South Carolina

OBJECTIVES

Ending the HIV epidemic requires innovative use of data for intelligent decision-making from surveillance through treatment. This study sought to examine the usefulness of using linked integrated PLWH health data to predict PLWH's future HIV care status and compare the performance of machine-learning methods for predicting future HIV care status for SC PLWH.

DESIGN

We employed supervised machine learning for its ability to predict PLWH's future care status by synthesizing and learning from PLWH's existing health data. This method is appropriate for the nature of integrated PLWH data because of its high volume and dimensionality.

METHODS

A data set of 8888 distinct PLWH's health records were retrieved from an integrated PLWH data repository. We experimented and scored seven representative machine-learning models including Bayesian Network, Automated Neural Network, Support Vector Machine, Logistic Regression, LASSO, Decision Trees and Random Forest to best predict PLWH's care status. We further identified principal factors that can predict the retention-in-care based on the champion model.

RESULTS

Bayesian Network (F = 0.87, AUC = 0.94, precision = 0.87, recall = 0.86) was the best predictive model, followed by Random Forest (F = 0.78, AUC = 0.81, precision = 0.72, recall = 0.85), Decision Tree (F = 0.76, AUC = 0.75, precision = 0.70, recall = 0.82) and Neural Network (cluster) (F = 0.75, AUC = 0.71, precision = 0.69, recall = 0.81).

CONCLUSION

These algorithmic applications of Bayesian Networks and other machine-learning algorithms hold promise for predicting future HIV care status at the individual level. Prediction of future care patterns for SC PLWH can help optimize health service resources for effective interventions. Predictions can also help improve retention across the HIV continuum.

Relationship between Health Literacy and Medication Adherence of Turkish Cancer Patients Receiving Oral Chemotherapy

Objective:

The aim of this study was to explore the relationship between health literacy and self-report medication adherence of Turkish cancer patients receiving oral chemotherapy.

Methods:

The present research was a descriptive and cross-sectional study and conducted with 100 voluntary cancer patients who were admitted to the medical oncology outpatient clinic and received oral chemotherapy. The data were collected through a questionnaire form consisting of the Oral Chemotherapy Adherence Scale and the Turkish Health Literacy Scale (TSOY-32). The collected data were analyzed using descriptive statistics, one-way ANOVA, and Pearson's correlation coefficient.

Results:

The results revealed that 57% of the patients were female, 35% were primary school graduates, 51% were breast cancer, and 36% took capecitabine. The mean index scores of the participants on both scales were calculated as 12.39 ± 1.51 and 73.25 ± 6.18, respectively. Overall, a positive and strong correlation was found between oral chemotherapy adherence and health literacy of the participants (r = 0.707, P = 0.000).

Conclusions:

Medication adherence and health literacy levels among the cancer patients in Turkey are alarming so that patient-centered interventions and training are required to overcome the barriers to medication adherence.

Association of Medication Adherence with Hospital Utilization and Costs Among Elderly with Diabetes Enrolled in a State Pharmaceutical Assistance Program

BACKGROUND

Medication adherence is crucial for the successful treatment among elderly patients with diabetes taking oral antidiabetic medications (OAMs). Cost of medications, lack of insurance coverage, and low income are major contributing factors towards medication nonadherence. State pharmaceutical assistance programs (SPAPs) provide medications at little or no cost to income-eligible patients and have potential to improve medication adherence among elderly patients. Despite this, limited research has focused on the association of medication adherence with health care utilization among elderly patients enrolled in SPAPs, and inclusion of health care costs as an outcome is even rarer.

OBJECTIVE

To evaluate the relationship between adherence to OAMs and hospital utilization and costs among elderly patients with diabetes who were enrolled in a SPAP.

METHODS

This retrospective observational study included elderly patients with diabetes enrolled in Pennsylvania's Pharmaceutical Assistance Contract for the Elderly (PACE) program in 2015. Medication adherence was estimated as the proportion of days covered (PDC; adherent: PDC≥80%, nonadherent: PDC < 80%). Hospital utilization and costs were estimated using hospital discharge records from the Pennsylvania Health Care Cost Containment Council. Multiple adjusted regression analyses were used to examine the association of medication adherence with hospital utilization (all-cause and diabetes-related number of inpatient hospital visits and length of stay [LOS]) and costs.

RESULTS

Among 9,497 elderly PACE enrollees with diabetes, 81% were adherent, and 21% were hospitalized. Compared with adherent patients, patients who were nonadherent to OAMs had twice the odds of all-cause and diabetes-related hospitalization. Controlling for covariates, nonadherent patients had 27% more all-cause (95% CI = 9%-36%) and 21% more diabetes-related (95% CI = 5%-40%) hospital visits than adherent patients. Covariate-adjusted LOS for nonadherent patients was 24% longer than that of adherent patients for all-cause hospitalization (95% CI = 1.171-1.311) and 12.7% longer for diabetes-related hospitalization (95% CI = 1.036-1.227). Medication nonadherence was associated with significantly greater all-cause ($22,670 vs. $16,383; P < 0.0001) and diabetes-related ($13,518 vs. $12,634; P = 0.0003) hospitalization costs.

CONCLUSIONS

Among SPAP-enrolled elderly patients, nonadherence to OAMs was significantly associated with increased risk of hospitalization, longer hospital stays, and greater hospitalization costs. Attention is needed to improve medication adherence among elderly receiving financial assistance to pay their prescriptions to reduce economic burden on the health care system.

DISCLOSURES

No outside funding supported this study. The authors have nothing to disclose.

Determining hypertensive patients' beliefs towards medication and associations with medication adherence using machine learning methods

Background

This study assesses the feasibility of using machine learning methods such as Random Forests (RF), Artificial Neural Networks (ANN), Support Vector Regression (SVR) and Self-Organizing Feature Maps (SOM) to identify and determine factors associated with hypertensive patients' adherence levels. Hypertension is the medical term for systolic and diastolic blood pressure higher than 140/90 mmHg. A conventional medication adherence scale was used to identify patients' adherence to their prescribed medication. Using machine learning applications to predict precise numeric adherence scores in hypertensive patients has not yet been reported in the literature.

Methods

Data from 160 hypertensive patients from a tertiary hospital in Kuala Lumpur, Malaysia, were used in this study. Variables were ranked based on their significance to adherence levels using the RF variable importance method. The backward elimination method was then performed using RF to obtain the variables significantly associated with the patients' adherence levels. RF, SVR and ANN models were developed to predict adherence using the identified significant variables. Visualizations of the relationships between hypertensive patients' adherence levels and variables were generated using SOM.

Result

Machine learning models constructed using the selected variables reported RMSE values of 1.42 for ANN, 1.53 for RF, and 1.55 for SVR. The accuracy of the dichotomised scores, calculated based on a percentage of correctly identified adherence values, was used as an additional model performance measure, resulting in accuracies of 65% (ANN), 78% (RF) and 79% (SVR), respectively. The Wilcoxon signed ranked test reported that there was no significant difference between the predictions of the machine learning models and the actual scores. The significant variables identified from the RF variable importance method were educational level, marital status, General Overuse, monthly income, and Specific Concern.

Conclusion

This study suggests an effective alternative to conventional methods in identifying the key variables to understand hypertensive patients' adherence levels. This can be used as a tool to educate patients on the importance of medication in managing hypertension.

Big Data Analytics and Sensor-Enhanced Activity Management to Improve Effectiveness and Efficiency of Outpatient Medical Rehabilitation

Numerous societal trends are compelling a transition from inpatient to outpatient venues of care for medical rehabilitation. While there are advantages to outpatient rehabilitation (e.g., lower cost, more relevant to home and community function), there are also challenges including lack of information about how patient progress observed in the outpatient clinic translates into improved functional performance at home. At present, outpatient providers must rely on patient-reported information about functional progress (or lack thereof) at home and in the community. Information and communication technologies (ICT) offer another option—data collected about the patient’s adherence, performance and progress made on home exercises could be used to help guide course corrections between clinic visits, enhancing effectiveness and efficiency of outpatient care. In this article, we describe our efforts to explore use of sensor-enhanced home exercise and big data analytics in medical rehabilitation. The goal of this work is to demonstrate how sensor-enhanced exercise can improve rehabilitation outcomes for patients with significant neurological impairment (e.g., from stroke, traumatic brain injury, and spinal cord injury). We provide an overview of big data analysis and explain how it may be used to optimize outpatient rehabilitation, creating a more efficient model of care. We describe our planned development efforts to build advanced analytic tools to guide home-based rehabilitation and our proposed randomized trial to evaluate effectiveness and implementation of this approach.

Cost-Related Medication Nonadherence Among Older Adults: Findings From a Nationally Representative Sample

OBJECTIVES

To estimate the rate of and risk factors associated with cost-related medication nonadherence among older adults.

DESIGN

Cross-sectional analysis of the 2017 National Health Interview Survey (NHIS).

SETTING

Nationally representative health interview survey in the United States.

PARTICIPANTS

Survey respondents, aged 65 years or older (n = 5701 unweighted) in the 2017 wave of the NHIS.

MEASUREMENTS

Self-reported, cost-related medication nonadherence (due to cost: skip dose, reduce dose, or delay or not fill a prescription) and actions taken due to cost-related medication nonadherence (ask for lower-cost prescription, use alternative therapy, or buy medications from another country) were quantified. We used a series of multivariable logistic regression analyses to identify factors associated with cost-related medication nonadherence. We also reported analyses by chronic disease subgroups.

RESULTS

In 2017, 408 (6.8%) of 5901 older adults, representative of 2.7 million older adults nationally, reported cost-related medication nonadherence. Among those with cost-related medication nonadherence, 44.2% asked a physician for lower-cost medications, 11.5% used alternative therapies, and 5.3% bought prescription drugs outside the United States to save money. Correlates independently associated with a higher likelihood of cost-related medication nonadherence included: younger age, female sex, lower socioeconomic levels (eg, low income and uninsured), mental distress, functional limitations, multimorbidities, and obesity (P < .05 for all). Similar patterns were found in subgroup analyses.

CONCLUSION

Cost-related medication nonadherence among older adults is increasingly common, with several potentially modifiable risk factors identified. Interventions, such as medication therapy management, may be needed to reduce cost-related medication nonadherence in older adults. J Am Geriatr Soc 67:2463-2473, 2019.

Using predictive analytics and big data to optimize pharmaceutical outcomes

PURPOSE

The steps involved, the resources needed, and the challenges associated with applying predictive analytics in healthcare are described, with a review of successful applications of predictive analytics in implementing population health management interventions that target medication-related patient outcomes.

SUMMARY

In healthcare, the term big data typically refers to large quantities of electronic health record, administrative claims, and clinical trial data as well as data collected from smartphone applications, wearable devices, social media, and personal genomics services; predictive analytics refers to innovative methods of analysis developed to overcome challenges associated with big data, including a variety of statistical techniques ranging from predictive modeling to machine learning to data mining. Predictive analytics using big data have been applied successfully in several areas of medication management, such as in the identification of complex patients or those at highest risk for medication noncompliance or adverse effects. Because predictive analytics can be used in predicting different outcomes, they can provide pharmacists with a better understanding of the risks for specific medication-related problems that each patient faces. This information will enable pharmacists to deliver interventions tailored to patients' needs. In order to take full advantage of these benefits, however, clinicians will have to understand the basics of big data and predictive analytics.

CONCLUSION

Predictive analytics that leverage big data will become an indispensable tool for clinicians in mapping interventions and improving patient outcomes.

Medication Adherence and Its Association with Health Literacy and Performance in Activities of Daily Livings among Elderly Hypertensive Patients in Islamabad, Pakistan

Background and Objective: Medication non-adherence is a preventable reason for treatment failure, poor blood pressure control among hypertensive patients and the geriatric population owing to poor physical activity is more vulnerable strata. The objective of this study is to investigate medication adherence and its associated factors among Pakistani geriatric hypertensive patients. Methods: A cross-sectional survey-based study was conducted at the out-patient department of the cardiac center from May 2018 to August 2018. A universal sampling technique was used to approach patients and 262 eligible consented patients were interviewed to collect information about socio-demographics, health, and disease-related characteristics using a structured questionnaire. The Morisky Levine Green test was used for the assessment of medication adherence. The Barthel index and single item literacy screener (SILS) was used to measure performance in activities of daily living and health literacy respectively. Chi-square tests and multivariate binary logistic regression analysis were performed to find factors by using SPSS version 20. Results: Of the total 262 participants, about 38.9% (n = 102) were scored 4 and considered adherent while 61.1% (n = 160) were considered as non-adherent. In logistic regression analysis, self-reported moderate (OR = 3.538, p = 0.009) and good subjective health (OR = 4.249, p = 0.008), adequate health literacy (OR = 3.369, p < 0.001) and independence in performing activities of daily living (OR = 2.968, p = 0.002) were found to be independent predictors of medication adherence among older hypertensive patients. Conclusion: Medication adherence among the older hypertensive population in Pakistan is alarmingly low. This clearly requires patient-centered interventions to overcome barriers and educating them about the importance of adherence.

Development and validation of a predictive model for the diagnosis of solid solitary pulmonary nodules using data mining methods

Background

The purpose of this study is to develop a predictive model to accurately predict the malignancy of solid solitary pulmonary nodule (SPN) by data mining methods.

Methods

A training cohort of 388 consecutive patients with solid SPNs was used to develop a predictive model to evaluate the malignancy of solid SPNs. By using SPSS Modeler, we utilized logistic regression (LR), artificial neural network (ANN), k-nearest neighbor (KNN), random forest (RF), and support vector machines (SVM) classifiers to build predictive models. Another cohort of 200 consecutive patients with solid SPNs was used to verify the accuracy of the predictive model. Predictive performance was evaluated using the area under the receiver operating characteristic curve (AUC).

Results

There was no significant difference in patients' characteristics between the training cohort and the validation cohort. The AUCs of LR, ANN, KNN, RF, and SVM models for the validation cohort were 0.874±0.0280 (P=0.605), 0.833±0.0351 (P=0.104), 0.792±0.0418 (P=0.014), 0.775±0.0400 (P=0.013), and 0.890±0.0323 (reference), respectively. The SVM algorithm had the highest AUC, and the best sensitivity (90.3%), specificity (80.4%), positive predictive value (93.9%), negative predictive value (71.2%) and accuracy (88.0%) for the validation cohort among the five models.

Conclusions

Data mining by SVM might be a useful auxiliary algorithm in predicting malignancy of solid SPNs.

Conceptual Model of Medication Adherence in Older Adults with High Blood Pressure-An Integrative Review of the Literature

BACKGROUND

Medication adherence (MA) is the most important controlling factor of high blood pressure (HBP). There are a few MA models, but they have not been successful in predicting MA completely. Thus, this study aimed to expand a conceptual model of MA based on an ecological approach.

METHODS

An integrative review of the literature based on theoretical and empirical studies was completed. Data source comprised: Medline (including PubMed and Ovid), ISI, Embase, Google scholar, and internal databases such as Magiran, Google, SID, and internal magazines. Primary English and Persian language studies were collected from 1940 to 2018. The steps of study included: (a) problem identification, (b) literature review and extracting studies, (c) appraising study quality, (d) gathering data, (e) data analysis using the directed content analysis, (f) concluding.

RESULTS

Thirty-six articles were finally included and analyzed. After analysis, predictors of MA in older adults with hypertension were categorized into personal, interpersonal, organizational, and social factors. Although the personal factors have the most predictors in sub-categories of behavioral, biological, psychological, knowledge, disease, and medication agents, social, organizational and interpersonal factors can have indirect and important effects on elderly MA.

CONCLUSION

There are many factors influencing MA of elderly with HBP. The personal factor has the most predictors. The designed model of MA because of covering all predictor factors, can be considered as a comprehensive MA model. It is suggested that future studies should select factors for study from all levels of the model.

Comparative analysis of predictive methods for early assessment of compliance with continuous positive airway pressure therapy

Background

Patients suffering obstructive sleep apnea are mainly treated with continuous positive airway pressure (CPAP). Although it is a highly effective treatment, compliance with this therapy is problematic to achieve with serious consequences for the patients’ health. Unfortunately, there is a clear lack of clinical analytical tools to support the early prediction of compliant patients.

Methods

This work intends to take a further step in this direction by building compliance classifiers with CPAP therapy at three different moments of the patient follow-up, before the therapy starts (baseline) and at months 1 and 3 after the baseline.

Results

Results of the clinical trial shows that month 3 was the time-point with the most accurate classifier reaching an f1-score of 87% and 84% in cross-validation and test. At month 1, performances were almost as high as in month 3 with 82% and 84% of f1-score. At baseline, where no information of patients’ CPAP use was given yet, the best classifier achieved 73% and 76% of f1-score in cross-validation and test set respectively. Subsequent analyzes carried out with the best classifiers of each time point revealed baseline factors (i.e. headaches, psychological symptoms, arterial hypertension and EuroQol visual analog scale) closely related to the prediction of compliance independently of the time-point. In addition, among the variables taken only during the follow-up of the patients, Epworth and the average nighttime hours were the most important to predict compliance with CPAP.

Conclusions

Best classifiers reported high performances after one month of treatment, being the third month when significant differences were achieved with respect to the baseline. Four baseline variables were reported relevant for the prediction of compliance with CPAP at each time-point. Two characteristics more were also highlighted for the prediction of compliance at months 1 and 3.

Trial registration

ClinicalTrials.gov Identifier, NCT03116958. Retrospectively registered on 17 April 2017.

Electronic supplementary material

The online version of this article (10.1186/s12911-018-0657-z) contains supplementary material, which is available to authorized users.

Adherence to Bisphosphonates among People Admitted to an Orthopaedic and Geriatric Ward at a University Hospital in Sweden

Oral bisphosphonates are the first choice of therapy to reduce the risk of osteoporotic fractures. These medications have generally poor oral bioavailability, which may further be reduced by concomitant intake of certain foods and drugs; therefore, it is vital to follow specific instructions. The aim with this study was to assess general adherence to oral bisphosphonates and adherence to specific administration instructions among people admitted to two wards at Umeå University hospital in Sweden. This interview study focuses on elderly patients living at home and prescribed oral bisphosphonates. Invited were 27 patients admitted to an orthopaedic ward and a geriatric ward during the period 28 March 2017 and 5 December 2017. In total, 21 patients were interviewed regarding their adherence to oral bisphosphonates. Out of 21 patients, 13 (62%) were considered non-adherent. The most common reason was calcium intake less than 2 h after oral administration of bisphosphonate (54%). The number of regularly prescribed drugs was significantly higher among patients rated non-adherent to bisphosphonates compared to those rated adherent (p = 0.004). Adherence to bisphosphonates administration instruction among elderly people living at home was limited. More research is needed to confirm these results and to investigate the reasons for non-adherence and how adherence to bisphosphonates can be improved.

Exploiting machine learning for predicting skeletal-related events in cancer patients with bone metastases

The aim of the bone metastases (BM) treatment is to prevent the occurrence of skeletal-related events (SREs). In clinical, physicians could only predict the occurrence of SREs by subjective experience. Machine learning (ML) could be used as predictive models in the medical field. But there is no published research using ML to predict SREs in cancer patients with BM. The purpose of this study was to assess the associations of clinical variables with the occurrence of SREs and to subsequently develop prediction models to help identify SREs risk groups.

We analyzed 1143 cancer patients with BM. We used the statistical package of SPSS and SPSS Modeler for data analysis and the development of the prediction model. We compared the performance of logistic regression (LR), decision tree (DT) and support vector machine(SVM). The results suggested that Visual Analog Scale (VAS) scale was a key factor to SREs in LR, DT and SVM model. Modifiable factors such as Frankel classification, Mirels score, Ca, aminoterminal propeptide of type I collagen (PINP) and bone-specific alkaline phosphatase (BALP) were identified. We found that the result of applying LR, DT and SVM classification accuracy was 79.2%, 85.8% and 88.2%, with 9, 4 and 8 variables, respectively.

In conclusion, DT and SVM achieved higher accuracies with smaller number of variables than the number of variables used in LR. ML techniques can be used to build model to predict SREs in cancer patients with BM.

Validation of the Korean Version of the Depression Coping Self-Efficacy Scale (DCSES-K)

Coping self-efficacy is regarded as an important indicator of the quality of life and well-being for community-dwelling patients with depression. The Depression Coping Self-Efficacy Scale (DCSES) was designed to measure self-efficacy beliefs related to the ability to perform tasks specific to coping with the symptoms of depression. The purpose of this study was to examine the psychometric properties of a Korean version of the Depression Coping Self-Efficacy Scale (DCSES-K) for community-dwelling patients with depression. A cross-sectional survey design was used. Content and semantic equivalence of the instrument using translation and back-translation of the DCSES was established. A convenience sample of 149 community-dwelling patients with depression was recruited from psychiatric outpatient clinics. The reliability alpha for the DCSES-K was .93, and the internal consistency was found to be acceptable. For convergent validity, DCSES-K score was positively correlated with the General Self-Efficacy Scale (GSES-K) score. For construct validity, significant differences in DCSES-K scores were found between a lower BDI group (mean=73.7, SD=16.54) and a higher BDI group (mean=53.74, SD=16.99) (t=7.19, p<.001). For the DCSES-K, 5 factors were extracted, accounting for 62.7% of the variance. Results of this study suggest that DCSES-K can be used as a reliable and valid measure for examining self-efficacy coping with depression for Korean community-dwelling patients with depression.

Medication prescription and adherence disparities in non valvular atrial fibrillation patients: an Italian portrait from the ARAPACIS study

Non-valvular atrial fibrillation (NVAF) represents a major health-care problem, needing an extensive and strict thrombosis prevention for stroke and cardiovascular (CV) disease risks. NVAF management guidelines recommend adequate antithrombotic and anti-atherosclerotic therapies. Medication adherence has been recognized as a pivotal element in health quality promotion and in the achievement of better clinical outcomes. We conducted a post-hoc analysis of the "Atrial fibrillation Registry for Ankle-brachial index Prevalence Assessment-Collaborative Italian Study (ARAPACIS)" with the aim of discerning differences in pharmacological management and medication adherence among NVAF Italian patients. Furthermore, data were analysed according to Italian geographical macro-regions (North, Center, South) to evaluate whether socioeconomic conditions might also influence medication adherence. Thus, we selected 1,366 NVAF patients that fulfilled the Morisky Medication Adherence Scale-4 items. Regional disparities in drug prescriptions were observed. In particular, in high-risk patients (CHA2DS2-VASc ≥2) oral anticoagulants were more prescribed in Northern and Center patients (61 and 60 %, respectively) compared to 53 % of high-risk Southern patients. Also, medication adherence showed a progressive decrease from North to South (78 vs. 60 %, p < 0.001). This disparity was independent of the number of drugs consumed for any reason, since prevalence of poly-therapy among the three macro-regions was similar. Our results show regional differences in NVAF patients' antithrombotic management and medication adherence, potentially reflecting well-known disparities in socioeconomic status among Italian regions. Future interventions promoting campaigns to global health-care education may be desirable to improve clinical outcomes in NVAF patients.

Association between skin conditions and depressive disorders in community-dwelling older adults

BACKGROUND

Depression is frequently observed in dermatologic patients. However, the association between depressive disorders and skin conditions has rarely been explored through population-based studies, especially within older-adult populations.

OBJECTIVE

To test this association in a representative sample of an older-adult population.

METHODS

Data came from the Survey on the Health of the Elderly (Enquête sur la Santé des Aìnés [ESA]), a longitudinal survey conducted in Quebec among 2,811 older adults. Cross-lagged panel models were used to simultaneously examine cross-sectional and longitudinal relationships between the presence of skin conditions and depressive disorders.

RESULTS

The prevalence of skin conditions was 13%, and the prevalence of depressive disorders among participants presenting with skin conditions was 11%. Our results indicated significant cross-sectional correlation (ζ  =  0.20) between skin conditions and depressive disorders, but no longitudinal association was observed.

CONCLUSION

Our results reinforce the hypothesis that skin conditions and depressive disorders are concurrently associated in older adults. However, no evidence of the predictive effect of skin problems on depression (and vice versa) was found in our community sample. Despite the deleterious effect of the coexistence of these problems in older adults, studies are lacking. This article highlights the importance of this issue and emphasizes the need for further research on this topic.

Korean Anaphora Recognition System to Develop Healthcare Dialogue-Type Agent

Objectives

Anaphora recognition is a process to identify exactly which noun has been used previously and relates to a pronoun that is included in a specific sentence later. Therefore, anaphora recognition is an essential element of a dialogue agent system. In the current study, all the merits of rule-based, machine learning-based, semantic-based anaphora recognition systems were combined to design and realize a new hybrid-type anaphora recognition system with an optimum capacity.

Methods

Anaphora recognition rules were encoded on the basis of the internal traits of referred expressions and adjacent contexts to realize a rule-based system and to serve as a baseline. A semantic database, related to predicate instances of sentences including referred expressions, was constructed to identify semantic co-relationships between the referent candidates (to which semantic tags were attached) and the semantic information of predicates. This approach would upgrade the anaphora recognition system by reducing the number of referent candidates. Additionally, to realize a machine learning-based system, an anaphora recognition model was developed on the basis of training data, which indicated referred expressions and referents. The three methods were further combined to develop a new single hybrid-based anaphora recognition system.

Results

The precision rate of the rule-based systems was 54.9%. However, the precision rate of the hybrid-based system was 63.7%, proving it to be the most efficient method.

Conclusions

The hybrid-based method, developed by the combination of rule-based and machine learning-based methods, represents a new system with enhanced functional capabilities as compared to other pre-existing individual methods.

Personal health records: a randomized trial of effects on elder medication safety

PURPOSE

To examine the impact of a personal health record (PHR) on medication-use safety among older adults.

BACKGROUND

Online PHRs have potential as tools to manage health information. We know little about how to make PHRs accessible for older adults and what effects this will have.

METHODS

A PHR was designed and pretested with older adults and tested in a 6-month randomized controlled trial. After completing mailed baseline questionnaires, eligible computer users aged 65 and over were randomized 3:1 to be given access to a PHR (n=802) or serve as a standard care control group (n=273). Follow-up questionnaires measured change from baseline medication use, medication reconciliation behaviors, and medication management problems.

RESULTS

Older adults were interested in keeping track of their health and medication information. A majority (55.2%) logged into the PHR and used it, but only 16.1% used it frequently. At follow-up, those randomized to the PHR group were significantly less likely to use multiple non-steroidal anti-inflammatory drugs-the most common warning generated by the system (viewed by 23% of participants). Compared with low/non-users, high users reported significantly more changes in medication use and improved medication reconciliation behaviors, and recognized significantly more side effects, but there was no difference in use of inappropriate medications or adherence measures.

CONCLUSIONS

PHRs can engage older adults for better medication self-management; however, features that motivate continued use will be needed. Longer-term studies of continued users will be required to evaluate the impact of these changes in behavior on patient health outcomes.

Real-data comparison of data mining methods in prediction of diabetes in iran

OBJECTIVES

Diabetes is one of the most common non-communicable diseases in developing countries. Early screening and diagnosis play an important role in effective prevention strategies. This study compared two traditional classification methods (logistic regression and Fisher linear discriminant analysis) and four machine-learning classifiers (neural networks, support vector machines, fuzzy c-mean, and random forests) to classify persons with and without diabetes.

METHODS

The data set used in this study included 6,500 subjects from the Iranian national non-communicable diseases risk factors surveillance obtained through a cross-sectional survey. The obtained sample was based on cluster sampling of the Iran population which was conducted in 2005-2009 to assess the prevalence of major non-communicable disease risk factors. Ten risk factors that are commonly associated with diabetes were selected to compare the performance of six classifiers in terms of sensitivity, specificity, total accuracy, and area under the receiver operating characteristic (ROC) curve criteria.

RESULTS

Support vector machines showed the highest total accuracy (0.986) as well as area under the ROC (0.979). Also, this method showed high specificity (1.000) and sensitivity (0.820). All other methods produced total accuracy of more than 85%, but for all methods, the sensitivity values were very low (less than 0.350).

CONCLUSIONS

The results of this study indicate that, in terms of sensitivity, specificity, and overall classification accuracy, the support vector machine model ranks first among all the classifiers tested in the prediction of diabetes. Therefore, this approach is a promising classifier for predicting diabetes, and it should be further investigated for the prediction of other diseases.