Pharmaceutical algorithms set in a real time clinical decision support targeting high-alert medications applied to pharmaceutical analysis

[Use of a pharmaceutical decision support system in the valuation of hospital stays: Evaluation through 3 examples in collaboration with the department of medical information]

Pharmacy decision support systems (PDSS) help clinical pharmacists to prevent and detect adverse drug events. The coding of hospital stays by the department of medical information (DMI) requires expertise, as it determines hospital revenues and the epidemiological data transmitted via the French national hospital database. The aim was to study the interest and feasibility of using a PDSS, in collaboration with the DMI, to help with the coding of hospital stays. Over 5 months, three rules were implemented in the PDSS to detect gout, Parkinson's disease and oro-pharyngeal candidiasis. The PDSS alerts were analyzed by a pharmacy resident and then forwarded to the DMI, who analyzed the stays to see whether or not the coding for the disease corresponding to the alert was present. The absence of coding was evaluated and tracked, along with the resulting change in severity and valuation. Three hundred and ninety-nine alerts from the PDSS were analyzed and sent to the DMI, representing 211 stays and 309 uniform hospital standardized discharge abstract (UHSDA) in the fields of medicine, surgery and obstetrics. Two hundred and eight (67.3%) UHSDA did not have the coding corresponding to the alert. For the majority of these UHSDAs, apart from diagnostic precision, there was no impact on the valuation of stays. For 4 UHSDAs, the addition of the diagnosis code led to an increase in the value of the stay and the severity of the homogeneous patient groups. The total revaluation corresponding to this modification was €5416. The use of PDSS has helped in the precision of diagnosis coding and the valuation of stays. This result must be weighed against the time invested in analyzing alerts and associated coding. An improvement in disease detection and data processing is needed to be feasible in practice, given the more than 227,600 RSS performed per year at our facility.

Clinical risk assessment of modelled situations in a pharmaceutical decision support system: a modified e-Delphi exploratory study

BACKGROUND

Pharmaceutical decision support systems (PDSSs) use reasoning software to match patient data to modelled situations likely to cause drug-related problems (DRPs) or adverse drug events. To aid decision-making, modelled situations must be linked to well-defined systemic clinical risks.

AIM

To obtain expert consensus on the level of clinical risk for patients associated with each modelled situation that could be addressed using a PDSS.

METHOD

A two-round e-Delphi survey was conducted from February to April 2022, involving 20 experts from four French-speaking countries. Participants had to rate modelled situations on two five-point Likert scales, assessing the likelihood of clinical consequences and their severity. The degree of consensus was determined as the proportion of participants providing risk scores in line with the median. The combined median scores for likelihood and severity provided the level of risk according to the Clinical Risk Situation for Patients (CRiSP) scale, formalized via validated tools.

RESULTS

The expert panel achieved consensus (≥ 75% agreement) on 48 out of 52 modelled clinical situations. Among these, 45 were categorized as high or extreme risk. The most common DRP identified was overdosing, accounting for 22% of cases. Furthermore, DRPs involving cardiovascular, psychiatric, and endocrinological drug classes were prevalent, constituting 45, 13, and 9% of cases, respectively.

CONCLUSION

Through consensus, our study identified 45 modelled clinical situations associated with high or extreme risks. This study highlights the interest of using PDSSs to prevent harm in patients and, on a large scale, document the impact of the pharmacist in preventing, intercepting and managing iatrogenic drug risk.

Deep learning application to automated classification of recommendations made by hospital pharmacists during medication prescription review

PURPOSE

Recommendations to improve therapeutics are proposals made by pharmacists during the prescription review process to address suboptimal use of medicines. Recommendations are generated daily as text documents but are rarely reused beyond their primary use to alert prescribers and caregivers. If recommendation data were easier to summarize, they could be used retrospectively to improve safeguards for better prescribing. The objective of this work was to train a deep learning algorithm for automated recommendation classification to valorize the large amount of recommendation data.

METHODS

The study was conducted in a French university hospital, at which recommendation data were collected throughout 2017. Data from the first 6 months of 2017 were labeled by 2 pharmacists who assigned recommendations to 1 of the 29 possible classes of the French Society of Clinical Pharmacy classification. A deep neural network classifier was trained to predict the class of recommendations.

RESULTS

In total, 27,699 labeled recommendations from the first half of 2017 were used to train and evaluate a classifier. The prediction accuracy calculated on a validation dataset was 78.0%. We also predicted classes for unlabeled recommendations collected during the second half of 2017. Of the 4,460 predictions reviewed, 67 required correction. When these additional labeled data were concatenated with the original dataset and the neural network was retrained, accuracy reached 81.0%.

CONCLUSION

To facilitate analysis of recommendations, we have implemented an automated classification system using deep learning that achieves respectable performance. This tool can help to retrospectively highlight the clinical significance of daily medication reviews performed by hospital clinical pharmacists.

Enhancing pharmaceutical decision support system: evaluating antithrombotic-focused algorithms for addressing drug-related problems

OBJECTIVES

To evaluate the efficacy of integrating antithrombotic-focused pharmaceutical algorithms (PAs) into a pharmaceutical decision support system (PDSS) for detecting drug-related problems (DRPs) and facilitating pharmaceutical interventions.

METHODS

A set of 26 PAs (12.4%) out of a total of 210 were created to model patient situations involving antithrombotics, and their contributions were compared with the entire PDSS system.The observational prospective study was conducted between November 2019 and June 2023 in two health facilities with 1700 beds. Pharmacists, who followed a DRP resolution strategy to support human supervision, analysed alerts generated by these encoded PAs. They registered their interventions and the acceptance by physicians.

RESULTS

From 3290 alerts analysed targeting antithrombotics, the pharmacists issued 1170 interventions of which 676 (57.8%) were accepted by physicians. With the 184 other PAs, from 9484 alerts the pharmacists issued 3341 interventions of which 1785 were accepted (53.4%).Results indicate that the detection of DRPs related to antithrombotics usage represents a high proportion of those detected by the PDSS, highlighting the importance of incorporating tailored PA elements at the modelling stage.

CONCLUSIONS

The system evolves alongside the physiological changes associated to the patient situations, adapts the alerts and complements the current care. Therefore, we recommend that all PDSS should integrate specific algorithms targeting DRPs associated with antithrombotics to enhance pharmaceutical interventions and improve patient safety.

[First French-speaking days of users of decision support system in clinical pharmacy: Feedback and perspectives]

Clinical decision support systems (CDSS) are tools that have been used for several years by clinical pharmacy teams to support pharmaceutical analysis, with a perspective of contributing to the quality of care in collaboration with the other health care team members. These tools require both technical, logistical and human resources. The growing use of these systems in different establishments in France and in Europe gave birth to the idea of meeting to share our experiences. The days organized in Lille in September 2021 aimed at proposing a time of exchange and reflection on the use of these CDSS in clinical pharmacy. A first session was devoted to feedback from each establishment. These tools are essentially used to optimize pharmaceutical analysis and to secure patient medication management. This session outlined the clear advantages and common limitations of these CDSS. Two research projects were also presented to put the use of these tools into perspective. The second session of these days, in the form of workshops, addressed 4 themes that surround the implementation of CDSS: their usability, the legal aspect, the creation of rules and their possible valorization. Common problems were raised, the resolution of which requires close collaboration. This is a first step proposing a beginning of harmonization and sharing that should be deepened in order not to lose the dynamics created between the different centers. This event ended with the proposal to set up two working groups around these systems: the creation and structuring of rules for the detection of risk situations and the common valorization of the work.

Assessment of chemotherapy-related educational needs of colorectal cancer patients

AIM

Aim of our study was to evaluate cancer patients' knowledge about their chemotherapy regimens in order to assess educational needs of patients.

METHODS

Study was conducted on 58 colorectal carcinoma patients who were treated in an outpatient chemotherapy unit. These patients had received a 2-page information pamphlet about their chemotherapy treatments before the commencement of treatment. During the first interview with patients, pharmacist collected demographic data and evaluated patients' knowledge about their medications using a standardized questionnaire.

FINDINGS

Mean age of the patients was 59.6 ± 1.3 years; 65.5% were male. Majority (77.6%) of patients were graduates of primary school. Sixty-four percent of these had at least one comorbid disease. Median number of chemotherapy courses already received by patients was 4 (1-9). Fifty-nine percent reported that they did not receive any patient education and 43.1% reported that they did not receive any informative document. Twenty-nine percent of patients did not know what actions to take in case of nausea-vomiting; while 53.4% did not know how to react if their body temperature exceeded 38 °C and 25.9% had no idea about dietary necessities. About one-third of patients did not pay attention to oral care.

CONCLUSION

Our study showed that patients did not understand (or remember) the basic points about their chemotherapy sufficiently, but remembered the adverse effects they experienced occasionally. Pharmacists will have the chance to increase the level of knowledge of the patients receiving chemotherapy by providing patient education and follow-up.