DETECT-IP: a Clinical Decision Support System and Intelligent Procedures to Counter Some Adverse Drug Events in Older Hospital Patients (DETECT-IP:)

DETECT-IP: a Clinical Decision Support System and Intelligent Procedures to Counter Some Adverse Drug Events in Older Hospital Patients

Reduction of Acute Renal Failure and/or Hyperkaliemia Adverse Drug Events in Older Inpatients by Incorporating Specific Rules Into a Computerized Support System and Dedicated Procedures: a Randomized Trial.

OméDIT (Observatory of Medicines, Medical Devices and Therapeutic Innovations, Regional Agency of Sante Nord Pas-de-Calais

Current evidence shows that computerized decision support systems (CDSS) have shown to be insufficiently effective to prevent adverse drug reactions (ADRs) at large scale (e.g. whole hospital). Several barriers for successful implementation of CDSS have been identified: over-alerting, lack of specificity of rules, and physician interruption during prescription. The effectiveness of CDSS could be increased in two ways. Firstly, by creating rules that are more specific to a given adverse drug reaction: the current study focuses on acute renal failure and hyperkalemia (two serious and frequent ADR in older hospitalized patients). Secondly, by involving the pharmacist in the review of the alerts so that he/she can transmit, if deemed necessary, a pharmaceutical recommendation to the clinician. This procedure will reduce over-alerting and prevent task interruption. The hypothesis is that the use of specific rules created by a multidisciplinary team and implemented in a CDSS, combined with a strategy for managing and transmitting alerts, can reduce specific ADRs such as hyperkalemia and acute renal failure.

Acute renal failure, hyperkaliemia, older people, computerized decision support system, stepped-wedge

Prospective, multicentre, controlled, single-blind, randomised cluster study with stepped-wedge permutations. The centres (hospitals) will be the clusters.

In the intervention group, the pharmaceutical validation will be based on routine care, often on entry to a ward and by analysis of all the alerts produced by the CDSS. Some alerts will result in a pharmaceutical intervention being provided to the medical team

In the control group, the pharmaceutical validation will be based on routine care, often on entry to a ward or in a particular situation

Number of adverse drug events such as acute renal failure and/or hyperkalemia in older hospitalized patients.

Presence of an adverse event related to the intervention provided ("change of prescription", "discontinuation of drug")

Therapeutic adaptations implemented in case of acute renal failure (ARF) or hyperkalemia upon hospital admission

Therapeutic changes within 72 hours of a CDSS alert for acute renal failure or hyperkalemia. Therapeutic changes include discontinuation of drug therapy, introduction of a new drug, dose reduction or change of drug

Relevance of CDSS alerts is defined in a standard way. Each CDSS alert is evaluated by a clinical pharmacist according to their own expertise and data available in the EHR. If the alert was deemed not relevant, the clinical pharmacist did not perform any pharmaceutical intervention. The CDSS software register the classification of the alert as "not relevant". This approach was used the last 4 years in our hospital and as been published in an article published in the International Journal of Medical Informatics: Cuvelier E, Robert L, Musy E, Rousselière C, Marcilly R, Gautier S, Odou P, Beuscart JB, Décaudin B. The clinical pharmacist's role in enhancing the relevance of a clinical decision support system. Int J Med Inform. 2021 Nov;155:104568. doi: 10.1016/j.ijmedinf.2021.104568. Epub 2021 Sep 2. PMID: 34537687

When an alert is received by the pharmacist, it is analyzed and the pharmacist forwards a pharmaceutical intervention to the physician in charge of the patient to propose a modification of the treatment (dosage, dose, stop

Changes in ADEs (Adverse Drug Event) prevention/management work process induced by the introduction of alerts

Changes in the work system are identified through a comparison of its elements (tools, tasks, organization, interactions, work environment, professionals), before and after the introduction of alerts, using qualitative system engineering methods.

Use medico-economic data such as time spent treating an alert, cost of treating an adverse drug reaction to estimate the cost-effectiveness of the intervention

Inclusion Criteria: Hospitalized for 3 days or more in an MCO (medicine surgery obstetrics) department participating in the study Patient who gave oral consent to participate in the study Socially insured patient Exclusion Criteria: Patient discharged or died before D3 of hospitalization Patient in palliative care or end of life on entry to the service Person under legal protection (curatorship) Lack of coverage by the social security system, Failure to obtain oral consent to participate in the study