Validation of a score to identify inpatients at risk of a drug-related problem during a 4-year period

Development of machine-learning models using pharmacy inquiry database for predicting dose-related inquiries in a tertiary teaching hospital

BACKGROUND

Drug-related problems (DRPs) are a significant concern in healthcare. Pharmacists play a vital role in detecting and resolving DRPs to improve patient safety. A pharmacy inquiry program was established in a tertiary teaching hospital to document inquiries about physicians' orders, aimed at preventing potential DRPs or providing medication information during order reviews.

OBJECTIVE

We aimed to develop machine-learning models using a pharmacy inquiry database to predict dose-related inquiries based on prescriptions and patient information.

METHODS

This retrospective study analyzed 20,393 pharmacy inquiries collected between January 2018 and February 2023. Data included prescription information (drug ingredient, dose, unit, and frequency), patient characteristics (age, sex, weight, and department), and renal function. The inquiries were categorized into two classes: dose-related inquiries (e.g., wrong dose and inappropriate regimen) and non-dose-related inquiries (e.g., inappropriate drug form and administration route). Six machine-learning models were developed: logistic regression, support vector classifier, decision tree, random forest, extreme gradient boosting, and categorical boosting. To evaluate the performance of the models, the area under the receiver operating characteristic curve and the accuracy were compared.

RESULTS

The CatBoost model achieved the highest performance (sensitivity: 0.92; accuracy: 0.79). The SHapley Additive exPlanations values highlighted the importance of features in the model predictions, drug ingredients, units, and renal function, in that order. Notably, lower renal function positively contributed to the prediction of dose-related inquiries. Additionally, the subsequent feature importance among drug ingredients showed that drugs such as acetylsalicylic acid, famotidine, metformin, and spironolactone strongly influenced the prediction of dose-related inquiries.

CONCLUSION

Machine-learning models that use pharmacy inquiry data can effectively predict dose-related inquiries. Further external validation and refinement of the models are required for broader applications in healthcare settings. These findings provide valuable guidance for healthcare professionals and highlight the potential of machine learning in pharmacists' decision-making.

Drug-related problems in patients admitted for SARS-CoV-2 infection during the COVID-19 pandemic

Introduction: Drug-related problems (DRP) are events or circumstances in which drug therapy does or could interfere with desired health outcomes. In December 2019, a new coronavirus, SARS-CoV-2, appeared. Little knowledge about this type of infection resulted in the administration of various drugs with limited use in other pathologies. Evidence about DRP in patients with COVID-19 is lacking. Objective: The aim of the present study is to describe identified cases of DRP and those drugs involved in the first wave of patients with COVID-19, and evaluate associated risk factors. Material and methods: Observational, retrospective study performed in a tertiary university hospital between 14th March 2020 and 31 May 2020 (corresponding to the first COVID-19 wave). We recruited patients admitted during the study period. Exclusion criteria included age < 18 years; admission to critically ill units; and care received either in the emergency room, at-home hospitalization or a healthcare center. Results: A total of 817 patients were included. The mean age was 62.5 years (SD 16.4) (range 18-97), and 453 (55.4%) were male. A total of 516 DRP were detected. Among the patients, 271 (33.2%) presented at least one DRP. The mean DRP per patient with an identified case was 1.9. The prevailing DRPs among those observed were: incorrect dosage (over or underdosage) in 145 patients (28.2%); wrong drug combination in 131 (25.5%); prescriptions not in adherence to the then COVID-19 treatment protocol in 73 (14.1%); prescription errors due to the wrong use of the computerized physician order entry in 47 (9.2%); and incorrect dosage due to renal function in 36 (7%). The logistic regression analysis showed that patients who received only prescriptions of antibacterials for systemic use (J01 ATC group) faced a higher likelihood of experiencing a DRP (OR 2.408 (1.071-5.411), p = 0.033). Conclusion: We identified several factors associated with an increased risk of DRPs, similar to those reported in other pre-pandemic studies, including a prolonged length of stay, higher number of prescribed drugs and antimicrobial administration. The relevance of pharmacists and tools like pharmacy warning systems can help prevent, identify and resolve DRP efficiently.

Systematic Review of Risk Factors Assessed in Predictive Scoring Tools for Drug-Related Problems in Inpatients

Drug-related problems (DRP, defined as adverse drug events/reactions and medication errors) are a common threat for patient safety. With the aim to aid improved allocation of specialist resources and to improve detection and prevention of DRP, numerous predictive scoring tools have been proposed. The external validation and evidence for the transferability of these tools still faces limitations. However, the proposed scoring tools include partly overlapping sets of similar factors, which may allow a new approach to estimate the external usability and validity of individual risk factors. Therefore, we conducted this systematic review and analysis. We identified 14 key studies that assessed 844 candidate risk factors for inclusion into predictive scoring tools. After consolidation to account for overlapping terminology and variable definitions, we assessed each risk factor in the number of studies it was assessed, and, if it was found to be a significant predictor of DRP, whether it was included in a final scoring tool. The latter included intake of ≥ 8 drugs, drugs of the Anatomical Therapeutic Chemical (ATC) class N, ≥1 comorbidity, an estimated glomerular filtration rate (eGFR) <30 mL/min and age ≥60 years. The methodological approach and the individual risk factors presented in this review may provide a new starting point for improved risk assessment.

Clinical usefulness of prediction tools to identify adult hospitalized patients at risk of drug-related problems: A systematic review of clinical prediction models and risk assessment tools

AIMS

This study aimed to review systematically all available prediction tools identifying adult hospitalized patients at risk of drug-related problems, and to synthesize the evidence on performance and clinical usefulness.

METHODS

PubMed, Scopus, Web of Science, Embase, and CINAHL databases were searched for relevant studies. Titles, abstracts and full-text studies were sequentially screened for inclusion by two independent reviewers. The Prediction Model Risk of Bias Assessment Tool (PROBAST) and the Revised Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2) checklists were used to assess risk of bias and applicability of prediction tools. A narrative synthesis was performed.

RESULTS

A total of 21 studies were included, 14 of which described the development of new prediction tools (four risk assessment tools and ten clinical prediction models) and six studies were validation based and one an impact study. There were variations in tool development processes, outcome measures and included predictors. Overall, tool performance had limitations in reporting and consistency, with the discriminatory ability based on area under the curve receiver operating characteristics (AUROC) ranging from poor to good (0.62-0.81), sensitivity and specificity ranging from 57.0% to 89.9% and 30.2% to 88.0%, respectively. The Medicines Optimisation Assessment tool and Assessment of Risk tool were prediction tools with the lowest risk of bias and low concern for applicability. Studies reporting external validation and impact on patient outcomes were scarce.

CONCLUSION

Most prediction tools have limitations in development and validation processes, as well as scarce evidence of clinical usefulness. Future studies should attempt to either refine currently available tools or apply a rigorous process capturing evidence of acceptance, usefulness, performance and outcomes.

Drug-related problems of antibiotic use in gastroenteritis related to patient therapy outcomes at Universitas Gadjah Mada Hospital

OBJECTIVES

Gastroenteritis is a disease of digestive system commonly occur among the people. Some cases of gastroenteritis are caused by bacteria, so it is treated by using antibiotics. Inappropriate use of antibiotics can be associated to Drug-Related Problems (DRPs). This study aims to identify patterns of potential DRPs of antibiotic use and analyze the effect of potential DRPs of antibiotic use toward the patient's therapeutic outcomes and length of stay.

METHODS

This is a retrospective cross-sectional study carried out by using patient's medical record. The study population was gastroenteritis patients at the inpatient ward of Universitas Gadjah Mada Hospital during January 2018-June 2019. Then, SPSS was employed to analyze the data and the effect of potential DRPs toward therapeutic outcomes was analyzed by utilizing the chi-square method.

RESULTS

More than half of gastroenteritis patients in Universitas Gadjah Mada Hospital were identified to have potential DRPs of antibiotic use. The most identified of potential DRPs was problems related to drug selection. Based on the chi-square analysis, there was no relationship between potential DRPs of antibiotic use and the therapeutic outcome. In addition, there was also no relationship between potential DRPs of antibiotic use and patient's length of stay.

CONCLUSIONS

The potential DRPs of antibiotics use do not have a significant effect on the therapeutic outcome and length of stay of the gastroenteritis patients in Universitas Gadjah Mada Hospital.

Drug related problems in clinical practice: a cross-sectional study on their prevalence, risk factors and associated pharmaceutical interventions

Drug-related problems (DRP) cause preventable negative health outcomes, especially during hospital admissions. The aim of our study was to examine the prevalence and characteristics of DRP in regular clinical pharmacy, as well as to determine those factors associated with a higher risk of DRP in the hospital setting. We analyzed data from a standardized registry database of regular pharmacy practice (2015- 2016). DRP were classified according to the Pharmaceutical Care Network Europe v6.2 classification. Cross-sectional data were obtained from 1602 adults admitted to medical wards. Crude and adjusted binary logistic regressions were performed to identify associations between potential risk factors and DRP. Overall DRP prevalence was high across medical specialties (45,1%), in a population characterized by advanced age, polypharmacy and multimorbidity. Problems leading to DRP were mainly classified into two domains (effectiveness and adverse reactions), being drug and dose selection the most frequent causes. Interventions were accepted and DRP were totally or partially solved in 74.1% and 4.81% of cases, respectively. In the adjusted model polypharmacy, allergies, BMI > 25 kg/m2 and clearance < 30 mL/min were associated with a higher risk of DRP. The participation of clinical pharmacists into multidisciplinary teams promotes the detection and solution of DRP. Polypharmacy, obesity, renal impairment and allergy are associated with a higher risk of DRP during admission.

Risk factors for drug-related problems in a general hospital: A large prospective cohort

OBJECTIVE

To identify risk factors for potential Drug-Related Problems (DRP) at admission in hospitalized patients.

METHODOLOGY

Prospective cohort study conducted in adults patients hospitalized (May 2016 to May 2018) in a general tertiary care hospital in Brazil. Potential DRP were detected by daily review of 100% of electronic medication orders by hospital pharmacists and classified by the Pharmaceutical Care Network Europe classification system (PCNE version 6.2). For the identification of risk factors of potential DRP, backward stepwise logistic regression was used to identify the set of independent predictors among over 120 variables collected in the initial 48 hours after admission in a training set consisting of 2/3 of the study population. The model was validated in the remaining sample.

RESULTS

The study population consisted of 1686 patients aged 52.0+/- 18.3 years-old, 51.4% females, with a median length of stay of 3.24 days, and 4.5% in-hospital mortality. The cumulative incidence of potential DRP was 14.5%. Admission for elective surgery and main diagnosis of disease of the circulatory system were associated with reduced risk of DRP (OR 0.41 and 0.57, respectively, p<0.05). The independent risk factors of DRP are heart rate ≥ 80 bpm (OR 1.41, p = 0.05), prescription of more than seven drugs in day 2 (OR 1.63, p = 0.05), prescription in day 1 of drugs of the Anatomical Therapeutic Chemical Code (ATC) class A (alimentary tract and metabolism, OR 2.24, p = 0.003), prescription in day 2 of two or more ATC class A drugs (OR = 3.52, p<0.001), and in day 1 of ATC class J drugs (antiinfectives for systemic use, OR 1.97, p = 0.001). In the validation set, the c-statistic of the predictive model was 0.65, the sensitivity was 56.1% and the specificity was 65.2%.

CONCLUSION

This study identified seven independent risk factors of potential DRP in patients hospitalized in a general hospital that have fair predictive performance for utilization in clinical practice.

Prevalence of drug-related problems using STOPP/START and medication reviews in elderly patients with dementia

BACKGROUND

Drug-related problems (DRPs) are common among elderly patients with dementia. STOPP/START is an explicit tool that has been used to detect DRPs among elderly patients.

OBJECTIVES

The objective of this study was to compare prevalence and type of DRPs identified by STOPP/START with DRPs identified by clinical pharmacists among the same population. Secondary objectives were to investigate factors associated with the use of DRPs using the two methods.

METHOD

Extracts from medical records were used to identify DRPs in 212 patients by using STOPP/START. The patients were ≥65 years of age with dementia or cognitive impairment. An earlier study was performed in the same study population in 2012-2014, where DRPs were identified by clinical pharmacists in order to decrease the number of rehospitalizations.

RESULTS

STOPP/START identified DRPs in 72.2% of the patients compared with 66.0% identified by the clinical pharmacists. The numbers of DRPs identified by the different methods were 326 and 310, respectively. Different types of DRPs were identified with the different tools. STOPP/START mainly identified DRPs in the categories "ineffective/inappropriate drug" and "needs additional drug therapy", whereas the clinical pharmacists identified DRPs in several categories.

CONCLUSION

Even though STOPP/START was able to identify a similar number of DRPs compared with DRPs identified by clinical pharmacists, STOPP/START failed to identify DRPs in several important categories. To cover all DRPs, STOPP/START might be used as a complement to implicit criteria.

Drug-related problems in patients with rheumatoid arthritis

BACKGROUND

Rheumatoid arthritis (RA) patients are at risk of acquiring drug-related problems (DRPs). However, there has been a lack of studies on DRPs in patients with RA up to now.

METHOD

This retrospective study was conducted in a tertiary hospital in Malaysia from January 2012 to December 2017 with the purpose of assessing DRPs in RA patients and factors associated with its occurrence. A total of 200 patients who had received pharmacological treatment for RA were enrolled in this study. Assessment of DRPs was based on the Pharmaceutical Network Care Europe tool version 5.01.

RESULTS

A total of 289 DRPs with an average of 1.5±1.0 problems per patient were identified, in which 78.5% of the population had at least one DRP. The most common DRPs encountered were adverse reactions (38.8%), drug interactions (33.6%), and drug-choice problems (14.5%). Factors that had significant association with the occurrence of DRPs were polypharmacy (P=0.003), multiple comorbidities (P=0.001), hyperlipidemia (P=0.009), osteo (P=0.040), and renal impairment (P=0.044). These data indicated that the prevalence of DRPs was high among RA patients.

CONCLUSION

Early identification of types of DRPs and associated factors may enhance the prevention and management of RA.