Psychometric Properties of a Machine Learning-Based Patient-Reported Outcome Measure on Medication Adherence: Single-Center, Cross-Sectional, Observational Study

BACKGROUND

Medication adherence plays a critical role in controlling the evolution of chronic disease, as low medication adherence may lead to worse health outcomes, higher mortality, and morbidity. Assessment of their patients' medication adherence by clinicians is essential for avoiding inappropriate therapeutic intensification, associated health care expenditures, and the inappropriate inclusion of patients in time- and resource-consuming educational interventions. In both research and clinical practices the most extensively used measures of medication adherence are patient-reported outcome measures (PROMs), because of their ability to capture subjective dimensions of nonadherence. Machine learning (ML), a subfield of artificial intelligence, uses computer algorithms that automatically improve through experience. In this context, ML tools could efficiently model the complexity of and interactions between multiple patient behaviors that lead to medication adherence.

OBJECTIVE

This study aimed to create and validate a PROM on medication adherence interpreted using an ML approach.

METHODS

This cross-sectional, single-center, observational study was carried out a French teaching hospital between 2021 and 2022. Eligible patients must have had at least 1 long-term treatment, medication adherence evaluation other than a questionnaire, the ability to read or understand French, an age older than 18 years, and provided their nonopposition. Included adults responded to an initial version of the PROM composed of 11 items, each item being presented using a 4-point Likert scale. The initial set of items was obtained using a Delphi consensus process. Patients were classified as poorly, moderately, or highly adherent based on the results of a medication adherence assessment standard used in the daily practice of each outpatient unit. An ML-derived decision tree was built by combining the medication adherence status and PROM responses. Sensitivity, specificity, positive and negative predictive values (NPVs), and global accuracy of the final 5-item PROM were evaluated.

RESULTS

We created an initial 11-item PROM with a 4-point Likert scale using the Delphi process. After item reduction, a decision tree derived from 218 patients including data obtained from the final 5-item PROM allowed patient classification into poorly, moderately, or highly adherent based on item responses. The psychometric properties were 78% (95% CI 40%-96%) sensitivity, 71% (95% CI 53%-85%) specificity, 41% (95% CI 19%-67%) positive predictive values, 93% (95% CI 74%-99%) NPV, and 70% (95% CI 55%-83%) accuracy.

CONCLUSIONS

We developed a medication adherence tool based on ML with an excellent NPV. This could allow prioritization processes to avoid referring highly adherent patients to time- and resource-consuming interventions. The decision tree can be easily implemented in computerized prescriber order-entry systems and digital tools in smartphones. External validation of this tool in a study including a larger number of patients with diseases associated with low medication adherence is required to confirm its use in analyzing and assessing the complexity of medication adherence.

Prostatic artery embolisation versus medical treatment in patients with benign prostatic hyperplasia (PARTEM): a randomised, multicentre, open-label, phase 3, superiority trial

Background

Prostatic artery embolisation (PAE) is a minimally invasive treatment of symptomatic benign prostatic hyperplasia (BPH). Our aim was to compare patient's symptoms improvement after PAE and medical treatment.

Methods

A randomised, open-label, superiority trial was set in 10 French hospitals. Patients with bothersome lower urinary tract symptoms (LUTS) defined by International Prostatic Symptom Score (IPSS) > 11 and quality of life (QoL) > 3, and BPH ≥50 ml resistant to alpha-blocker monotherapy were randomly assigned (1:1) to PAE or Combined Therapy ([CT], oral dutasteride 0.5 mg/tamsulosin hydrochloride 0.4 mg per day). Randomisation was stratified by centre, IPSS and prostate volume with a minimisation procedure. The primary outcome was the 9-month IPSS change. Primary and safety analysis were done according to the intention-to-treat (ITT) principle among patients with an evaluable primary outcome. ClinicalTrials.gov Identifier: NCT02869971.

Findings

Ninety patients were randomised from September 2016 to February 2020, and 44 and 43 patients assessed for primary endpoint in PAE and CT groups, respectively. The 9-month change of IPSS was -10.0 (95% confidence interval [CI]: -11.8 to -8.3) and -5.7 (95% CI: -7.5 to -3.8) in the PAE and CT groups, respectively. This reduction was significantly greater in the PAE group than in the CT group (-4.4 [95% CI: -6.9 to -1.9], p = 0.0008). The IIEF-15 score change was 8.2 (95% CI: 2.9-13.5) and -2.8 (95% CI: -8.4 to 2.8) in the PAE and CT groups, respectively. No treatment-related AE or hospitalisation was noticed. After 9 months, 5 and 18 patients had invasive prostate re-treatment in the PAE and CT group, respectively.

Interpretation

In patients with BPH ≥50 ml and bothersome LUTS resistant to alpha-blocker monotherapy, PAE provides more urinary and sexual symptoms benefit than CT up to 24 months.

Funding

French Ministry of Health and a complementary grant from Merit Medical.

Conditional prescriptions of oral antihypertensive drugs for the management of hypertension urgencies in the inpatient setting: An observational study

WHAT IS KNOWN AND OBJECTIVES

The management of hypertension urgencies during hospitalization may generally not necessitate urgent care. However, physicians frequently prescribe 'as needed' antihypertensive drugs for which administration is triggered by blood pressure thresholds. The lack of rationale for this hospital practice led us to study oral conditional antihypertensive (OCA) prescriptions. We aimed to estimate the prevalence of OCA prescriptions and to establish their characteristics.

METHODS

In our institution, prescriptions are computerized. The study was retrospectively performed using a hospital clinical data warehouse over a 5-year period.

RESULTS AND DISCUSSION

The prevalence of OCA prescriptions was 6.9% among subjects treated with an antihypertensive drug. The median duration of these prescriptions was 4 days, until the day of the patient discharge in 78.8% stays. The calcium channel inhibitors were the main (79.9%) pharmacological class prescribed, with mostly prescriptions of nicardipine. OCA prescriptions were associated with another antihypertensive medication in 58.8% of the prescriptions; for 19.3%, it was a medication belonging to the same pharmacological class than the OCA drug prescribed. Regarding the computerized drafting, 39.6% of the conditional prescriptions were considered uninterpretable. At least one administration by nurses concerned 65.1% of the OCA prescriptions. The mean SBP and DBP before the initiation of an OCA drug was 142.9 ± 28.2 and 75.8 ± 24.5 mm Hg, respectively, relative to 143.0 ± 24.9 and 77.6 ± 19.9 mm Hg after the initiation (P = .8 for SBP and P = .06 for DBP).

WHAT IS NEW AND CONCLUSION

The originality of this study lies in the use of a clinical data warehouse to evaluate OCA prescriptions in hospital. These prescriptions are current, often uninterpretable and mostly ordered until patient discharge. Such drug orders could be associated with an increased risk of iatrogenic events and/or administration errors. This underlies the need for developing decision support tools and computerized protocols to manage hypertension urgencies.

Risk scores identifying elderly inpatients at risk of 30-day unplanned readmission and accident and emergency department visit: a systematic review

OBJECTIVES

The aim of this systematic review was to describe and analyse the performance statistics of validated risk scores identifying elderly inpatients at risk of early unplanned readmission.

DATA SOURCES

We identified potentially eligible studies by searching MEDLINE, EMBASE, COCHRANE and Web of Science. Our search was restricted to original studies, between 1966 and 2018.

ELIGIBILITY CRITERIA

Original studies, which internally or externally validated the clinical scores of hospital readmissions in elderly inpatients.

DATA EXTRACTION AND SYNTHESIS

A data extraction grid based on Strengthening the Reporting of Observational Studies in Epidemiology and Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis statements was developed and completed by two reviewers to collect general data. The same process was used to extract metrological data of the selected scores.

QUALITY ASSESSMENT OF THE INCLUDED STUDIES

Assessment of the quality and risk of bias in individual studies was performed by two reviewers, using the validated Effective Public Health Practice Project quality assessment tool.

PARTICIPANTS

Elderly inpatients discharged to home from hospital or returning home after an accident and emergency department visit.

RESULTS

A total of 12 studies and five different scores were included in the review. The five scores present area under the receiving operating characteristic curve between 0.445 and 0.69. Identification of Senior At Risk (ISAR) and Triage  Risk Screening Tool (TRST) scores were the more frequently validated scores with ISAR being more sensitive and TRST more specific.

CONCLUSIONS

The TRST and ISAR scores have been extensively studied and validated. The choice of the most suitable score relies on available patient data, patient characteristics and the foreseen clinical care intervention. In order to pair the intervention with the appropriate clinical score, further studies of external validation of clinical scores, identifying elderly patients at risk of early unplanned readmission, are needed.

PROSPERO REGISTRATION NUMBER

CRD42017054516.

External validation of the 80+ score and comparison with three clinical scores identifying patients at least 75 years old at risk of unplanned readmission within 30 days after discharge

AIM OF THE STUDY

A potentially avoidable readmission is an unplanned readmission occurring within 30 days of discharge. As 20% of hospitalised elderly patients are rehospitalised as an unplanned readmission, it is necessary to identify with a clinical score those patients who are at risk of readmission and need discharge interventions as a priority. The main objective of this study was to externally validate and compare the 80+ score with the three other scores predicting the risk of unplanned readmission.

METHODS

We conducted a retrospective case-control study using a clinical data warehouse. The study included patients hospitalised between 1 September 2014 and 31 October 2015 in an 800-bed university hospital. We included patients aged 75 and over. Cases were readmitted at the emergency department within 30 days after the index discharge. Controls were not readmitted as an emergency within 30 days. Four clinical scores (80+ score, LACE index, HOSPITAL score, TRST) were externally validated. Discrimination of the scores was assessed by calculating the area under the receiver operating characteristic curves (AUC). Calibration was assessed with a Hosmer-Lemeshow χ2 test.

RESULTS

We included 438 patients. For discrimination, the 80+ score, the LACE index, the HOSPITAL score and the TRST had AUCs of 0.506 (95% confidence interval [CI] 0.413-0.546), 0.534 (95% CI 0.459-0.591, 0.517 (95% CI 0.466-0.598) and 0.589 (95% CI 0.524-0.654), respectively. The Hosmer-Lemeshow χ2 tests had p-values of 0.44, 0.43, 0.11 and 0.49, respectively.

CONCLUSION

In our study, the 80+ score was externally validated and showed less favourable discrimination than the three other scores in this population.

Iatrogenic risk factors associated with hospital readmission of elderly patients: A matched case-control study using a clinical data warehouse

WHAT IS KNOWN

Hospital readmission within 30 days of patient discharge has become a standard to judge the quality of hospitalizations. It is estimated that 14% of the elderly, people over 75 years old or those over 65 with comorbidities, are at risk of readmission, of which 23% are avoidable. It may be possible to identify elderly patients at risk of readmission and implement steps to reduce avoidable readmissions.

OBJECTIVE

The aim of this study was to identify iatrogenic risk factors for readmission. The secondary objective was to evaluate the rate of drug-related readmissions (DRRs) among all readmissions and compare it to the rate of readmissions for other reasons.

METHODS

We conducted a retrospective, matched, case-control study to identify non-demographic risk factors for avoidable readmission, specifically DRRs. The study included patients hospitalized between 1 September 2014 and 31 October 2015 in an 800-bed university hospital. We included patients aged 75 and over. Cases consisted of patients readmitted to the emergency department within 30 days of initial discharge. Controls did not return to the emergency department within 30 days. Cases and controls were matched on sex and age because they are known as readmissions risk factors. After comparison of the mean or percentage between cases and controls for each variable, we conducted a conditional logistic regression.

RESULTS

The risk factors identified were an emergency admission at the index hospitalization, returning home after discharge, a history of unplanned readmissions and prescription of nervous system drugs. Otherwise, 11.4% of the readmissions were DRRs, of which 30% were caused by an overdose of antihypertensive. The number of drugs at readmission was higher, and potentially inappropriate medications were more widely prescribed for DRRs than for readmissions for other reasons.

WHAT IS NEW AND CONCLUSION

In this matched case-control retrospective study, after controlling for gender and age, we identified the typical profile of elderly patients at risk of readmission. These patients had an unplanned admission at the index hospitalization and prescribed nervous system drugs at discharge from the index admission; they have a history of unplanned readmission within 30 days and return home after discharge.

A systematic review of regulatory and educational interventions to reduce the burden associated with the prescriptions of sedative-hypnotics in adults treated for sleep disorders

BACKGROUND

The burden of Sedative-Hypnotics (SHs) has been known since the 1980s. Yet, their consumption remains high. A systematic review of the literature should help to assess efficient interventions to improve the appropriate use of SHs in sleep disorders.

OBJECTIVES

To identify and assess regulatory and educational interventions designed to improve the appropriate use of SHs for insomnia treatment.

METHODS

We conducted a systematic review of the literature according to PRISMA guidelines. A systematic search covering the period 1980-2015 was carried out in Medline, Web of Science, Embase and PsycInfo. We included studies reporting the implementation of regulatory or educational strategies directed towards patients and/or healthcare professionals to improve the appropriate use of SHs to treat insomnia in the community, hospitals and nursing homes.

RESULTS

Thirty-one studies were included: 23 assessed educational interventions (recommendations by mail/email, computer alerts, meetings, mass media campaigns, prescription profile), 8 assessed regulatory interventions (prescription rule restriction, end of reimbursement). The most recent was implemented in 2009. Restrictive prescription rules were effective to reduce the consumption of targeted SHs but led to a switch to other non-recommended SHs. Among educational interventions, only 3 studies out of 7 reported positive results of mono-faceted interventions; whereas, 13 out of the 16 multi-faceted interventions were reported as efficient: particularly, the active involvement of healthcare professionals and patients and the spread of information through mass media were successful. The risk of bias was high for 24 studies (mainly due to the design), moderate for 3 studies and weak for 4 studies.

CONCLUSION

Educational multifaceted studies are presented as the most efficient. But further better designed studies are needed to make evidence-based results more generalizable.

Cardiovascular risk associated with high sodium-containing drugs: A systematic review

Background

Excess dietary sodium is associated with increased blood pressure (BP). Some drugs are associated with high sodium intake (in particular effervescent tablets), but the cardiovascular risk associated with such high sodium-containing drugs (HSCD) is largely underevaluated.

Objectives

To summarize the evidence for a potential cardiovascular risk associated with exposure to HSCD, and to highlight possible risk factors associated with this iatrogenic issue; in general and/or specific populations.

Methods

We conducted a systematic review, by searching electronic databases including MEDLINE, EMBASE, Web of Science, CENTRAL and grey literature between 1960 and 2015. We included studies that reported modification of cardiovascular parameters or incidence/prevalence of cardiovascular outcomes, between a group of subjects exposed to HSCD relative to a non-exposed group. The threshold used to identify HSCD was 391 mg/day. We did not consider studies evaluating exposure to sodium as an active ingredient or those focusing on dialysis solutions or enteral/parenteral nutrition. Study quality was assessed using the EPHPP tool.

Results

A total of eight studies met our inclusion criteria. Four reported results for short-term exposure to HSCD (≤ 7 days) on BP fluctuations. One study reported an elevation of BP (associated sodium intake: 1,656 mg/day). Four studies evaluated a long-term exposure (≥ 2 years or discontinuation of a chronic treatment). Two studies reported iatrogenic risk. For these studies, drug associated sodium intake was high (> 1,500 mg/day) in patients with comorbidities (in particular, diabetes mellitus and hypertension).

Conclusion

Despite numerous study limitations, this systematic review suggests three potential synergistic risk factors for cardiovascular complications after exposure to HSCD: a high sodium intake (≥ 1,500 mg/day), a long duration of exposure, and the presence of comorbidities. Further studies are required to characterize this iatrogenic risk.

Trial registration

PROSPERO CRD42016047086.

A simulation study of the interception of prescribing errors by clinical pharmacists in an acute hospital setting

RATIONALE, AIMS AND OBJECTIVES

To evaluate the performance of several pharmacists in the same department who analysed the same prescriptions in a simulation study.

METHODS

One hundred prescriptions were retrospectively extracted from the prospective database of our hospital. Five clinical pharmacists working in the same department were asked to analyse individually the order lines of each prescription as if it were part of their routine daily practice. Afterward, an independent committee of five other clinical pharmacists reviewed the same 100 prescriptions. We calculated the sensitivity and the specificity of error detection in a line order by using the results of the committee as the gold standard.

RESULTS

A total of 908 order lines were analysed (mean 9 ± 3 order lines per prescription). Fifty-one medication errors were identified by the committee (5.6%), including 23 related to laboratory test results: renal failure, or therapeutic concentrations being too low or too high. The sensitivity of the five pharmacists ranged between 19.6% and 56.9% and the specificity between 92.8% and 98.7%. The rates of agreement between each pharmacist and the committee, assessed using kappa coefficient, were between 0.20 and 0.39. The main factors affecting sensitivity and/or specificity in univariate analysis were the number of drugs per prescription, type of drug prescribed (ATC classification) and the glomerular filtration rate.

CONCLUSION

Discrepancies between the performances of pharmacists exist, as there are between other health care professionals. Pharmacist training, standardization of the pharmaceutical analysis of drug prescription, and implementation of a clinical decision support system allowing biological values to be linked to drug prescriptions could improve individual performance.

Information about medication in HIV-infected patients and its relation to adherence

QUESTIONS UNDER STUDY

In HIV-infected patients, comprehension of medication instructions is an essential condition for adherence to Highly Active Antiretroviral Therapy (HAART). In this study, we used a self-reported questionnaire to know which sources of medication information HIV-infected patients used and their impact on adherence. In secondary objectives, we determined profiles of non-adherent patients and specified the role of the pharmacist.

METHODS

A cross-sectional, observational study was conducted in one community pharmacy and one French university hospital pharmacy, in HAART-naïve or not patients, from April to June 2009.

RESULTS

During the 3-month study period, 233 HIV-infected patients were included. The majority of patients sought information about their HAART treatments from the hospital physician (79.8%), the community physician (74.2%), and patient information leaflets (73.8%). The community and hospital pharmacists were consulted by respectively 16.3% and 3.4% of patients. According to multivariate regression analysis, adherence seemed to be associated with the sources of information "community physician", "hospital physician", "internet", and the potential support of patient associations. A total of 65.7% of patients were considered to be adherent.

CONCLUSIONS

In our study, among sources used by HIV-infected outpatients, their physicians are the most helpful sources of information about HAART. Regarding practice implications, the key role of the pharmacist is underutilised, indicating the need for improved communication between the pharmacist and outpatients.

Knowledge of oral drug treatment in immunocompromised patients on hospital discharge

PRINCIPLES

Several studies have shown that patients' inappropriate knowledge about their medication is associated with non-adherence. The aim of this study was to assess immunocompromised inpatient knowledge of their oral drug treatment on discharge.

METHODS

We conducted a single-centre, prospective, cross-sectional study from July to November 2008 in the Immunology unit of a university-based hospital. Knowledge of all oral prescribed medication was assessed before discharge of immunocompromised inpatients using a self-administered questionnaire, assessing drug name, dosage, indication and administration guidelines. Prescribed drugs were classified as treatments for chronic disease, or as adjuvant treatments which were differentiated regarding their link with the chronic disease.

RESULTS

Over four months, 17 transplant recipients and 38 HIV-infected patients were included. Overall, 57% of the 497 prescribed drugs were adequately known. The proportions of drugs adequately known were 79%, 91%, 81% and 62% respectively for the drug name, dosage, indication and administration guideline components. Drugs for the treatment of chronic disease were more adequately known than adjuvant treatments. Older age and a low educational level were significantly associated with poor knowledge of drugs.

CONCLUSIONS

Immunocompromised patients demonstrated moderate to good knowledge of oral drugs on discharge. Adjuvant treatments were less well known than drugs for the treatment of chronic disease. Some recommendations for interventions aimed at utilising the skills of clinical pharmacists are needed. Efforts which encourage patients to be active participants in their own treatment could improve therapeutic adherence and reduce potential complications.

Non-adherence with drug treatment after heart or lung transplantation in adults: a systematic review

OBJECTIVE

Heart or lung transplantation is a complex intervention requiring medication adherence. The objective of this systematic review is to estimate the prevalence of non-adherence (NA) with post-transplantation medication in heart or lung recipients and to assess its clinical impact. We examined in the selected studies if the authors considered the patient's perspective in their evaluations.

METHODS

The electronic database MEDLINE, EMBASE and The Cochrane Central Register were searched. Only studies that reported the number of non-adhere subjects were eligible. The different methods of measurement, the ways in which authors defined NA and if authors had integrated patient's perspective in their secondary objectives were also assessed.

RESULTS

The range frequency of NA was 1-42.9% for all drugs. Non-adherent patients tend to experience worse outcomes compared to adherent patients. The patient's perception of drug side-effects is the most reported patient-related factor for impairing adherence.

CONCLUSION

NA after heart or lung transplantation is an important issue and concerns not only immunosuppressant treatments. The main striking point of the selected studies is the lack of patient perspective and the omission of patients-healthcare providers' relationship.

PRACTICE IMPLICATIONS

Future research must focus on patients' motivation for the medication-taking behaviour.