If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Recognition/reconciliation reduces inappropriate drug prescription in older medical patients.
A computerized aid program was thus adopted by hospital pharmacists and clinicians.
Many clinically significant drug-related problems were so identified.
The program led to deprescribing or closer drug monitoring.
In older medical patients polypharmacy is often associated with poor prescription appropriateness and harmful drug-drug interactions. An effort that jointly involved hospital pharmacists and clinicians attending multimorbid older patients acutely admitted to medical wards was implemented for medication recognition and reconciliation aided by the use of a computerized support system.
Six internal medicine wards enrolled consecutively 90 acutely admitted multimorbid patients aged 75 years or more taking 5 or more different drugs. Two hospital pharmacists carried out the recognition of medications taken at hospital ward admission, and interacted with the clinicians in a process of drug reconciliation, using also the computerized support system to evaluate drug related problems, prescription inappropriateness or drug-drug interactions. The process was repeated at hospital discharge.
Among a total number of 911 drugs prescribed to 90 older medical patients at ward admission, the pharmacists identified during their recognition/reconciliation 455 drug-related problems, mainly due to prescription of medications inappropriate for older multimorbid patients and clinically harmful drug-drug interactions. When these drug-related problems were identified by the pharmacist, the attending clinicians accepted and implemented the suggestions for changes for approximately two thirds of the discrepancies, thereby leading to deprescribing the implicated drugs or at least to their closer monitoring.
This interventional prospective study based upon the integrated expertise of hospital pharmacists and clinicians confirms that drug-related problems are frequent in multimorbid older patients acutely admitted to hospital medical wards, and demonstrates afresh the feasibility and mutual acceptance of a trajectory of recognition/reconciliation based upon an integrated collaboration between hospital pharmacists and ward clinicians in the process of medication optimization.
Older patients, the largest fraction of those acutely admitted to hospital medical wards, are often affected by multiple chronic diseases, that are in turn associated with the intake of multiple drugs [
REPOSI Collaborators (REPOSI is the acronym of REgistro POliterapie SIMI, Società Italiana di Medicina Interna) Polypharmacy in older people: lessons from 10 years of experience with the REPOSI register.
]. Another cause is the frequent fragmentation of prescriptions made by the different specialists often attending multimorbid older patients, with poor coordination and integration between them and generalist physicians such as internists, geriatricians and general practitioners in the implementation of an holistic strategy of medication review [
]. Furthermore, it is associated with negative outcomes, the most frequent and relevant being the adverse effects of the intake of drugs inappropriate for frail or prefrail older people, as well as clinically harmful drug-drug interactions (DDI) [
]. Avoidance of the resulting negative consequences warrants the implementation of a regular process of medication recognition and reconciliation, followed by dose adjustment or deprescribing of those drugs that were inappropriately prescribed and carry a risk of major adverse events in the elderly [
]. However, the process of recognition/reconciliation is often jeopardized by the relatively limited expertise on pharmacokinetics and pharmacodynamics by clinicians who prescribe medications in older patients, and by their scanty integration with the hospital pharmacists who dispense them [
], that supplies information not only on the existence and clinical relevance of potential DDIs but also on other aspects of poor prescription appropriateness. We chose to design a prospective interventional study carried out in the medical wards of an Italian research and teaching hospital, with the driving involvement of hospital pharmacists, who had the goal to identify and jointly evaluate with the attending clinicians: (1) the frequency of intentional (documented and not documented) and unintentional drug prescription discrepancies recorded in the frame of a process of medication recognition/reconciliation done at hospital admission and repeated at discharge, (2) the frequency and type of DRPs identified both at admission and discharge, (3) whether or not the suggestions made by the pharmacists in the frame of the medication recognition/reconciliation process were accepted by clinicians.
2. Patients and methods
This interventional prospective study was based upon the collaboration between the IRCCS Mario Negri Institute of Pharmacological Research and the IRCCS Maggiore Policlinico hospital, both in Milano, Italy. The hospital units involved were the Pharmacy and Scientific Direction and four Internal Medicine plus the Acute Medicine and Geriatric wards. The study protocol was approved by the Institutional Review Board and conducted according to Good Clinical Practice and the Declaration of Helsinki.
Each of the 6 participating clinical units was asked to consecutively enroll 15 multimorbid patients aged 75 years or more, provided they were admitted within the previous 72 h with an acute illness, were chronically prescribed at least 5 drugs and had a life expectancy of at least 6 months. When each patient with these features was admitted to the six medical wards and enrolled after obtaining their informed consent, one of two hospital pharmacists involved in the project and specially trained for this project personally contacted the attending clinicians in the medical wards in order to carry out recognition of the drugs taken by patients before and at hospital admission, that is, those prescribed at home by the general practitioners, in the hospital emergency room and ultimately in the medical wards after admission. This process was repeated before hospital discharge. The pharmacist-driven intervention was standardized and developed in three phases (Fig. 1): (1) drug recognition and production of a medication list (the information was obtained by interviewing patients/caregivers/family members as well as by inspecting hospital medical records). (2) medication reconciliation using as support INTERCheck, but also other criteria such as the Summary of Product Characteristics (SPC) of each drug and the national recommendations for appropriate drug prescription, provided in this instance by the Italian Medicine Agency (https://www.aifa.gov.it/note-aifa). During the recognition/reconciliation process, the pharmacists evaluated whether or not drug prescription was appropriate in older patients according to the most current versions of the BEERS, STOPP and START criteria [
] (Screening Tool of Older Persons’ Potentially inappropriate Prescription/Screening Tool to Alert doctor to the Right Treatment); whether drug dosing and intake duration needed adjustments depending on the liver and kidney functions in older people, and whether moderate to severe DDIs were detected according to INTERCheck. This computer tool, previously validated in the clinical setting, is currently available free of charge upon a motivated request as an online version, which provides the following information meant to optimize prescription appropriateness: potentially inappropriate medications in the elderly (according to the STOPP and BEERS criteria), drug-related anticholinergic burden, need of dosage changes in relation to renal or liver function, an algorithm that identifies patients at high risk of adverse drug reactions (ADR-Gerontonet Score) [
], the number needed to treat (NNT) and the number needed to harm (NNH) of the prescribed medications. Moreover, INTERCheck provides a list of potential DDIs classified in four levels of clinical relevance: D, drug combination that must be avoided owing to a very high risk of DRPs; C, major DDI, so that the combination requires at least close monitoring owing to such potentially severe clinical consequences as adverse effects or lack of clinical efficacy, B, moderate DDI, so that the combination requires dose adjustment and/or blood concentration monitoring; A, minor DDI, so that the combination is associated with interactions of no known clinical relevance. At the end of the recognition/reconciliation process the pharmacist(s) communicated to the attending clinicians the discrepancies identified and jointly classified them into three different categories:
intentional but not documented, that is, prescriptions made by clinicians on the basis of their evaluation of the patients’ current clinical status but without explicit motivation/documentation of the underlying reason(s),
intentional and documented, if made as above but with the provision of explicit motivations,
unintentional, if made unwillingly.
The pharmacists provided with the list of the DRPs identified during the recognition/reconciliation process the attending clinicians, who in turn had the opportunity to discuss with them information on the identified discrepancies and potential DRPs, and to choose whether or not the suggested changes were to be accepted and implemented. The same pharmacist-driven process was repeated as above at the time of hospital discharge and the prescription recommendations, together with the associated reasons, were communicated to the patients’ general practitioners by means of a structured letter.
3.1 General patient characteristic
Over the 12-month recruitment period the 6 participating medical wards screened 284 patients who met the age limit criterion of at least 75 years. However, only 90 of them (32%) were enrolled, the remaining ones being excluded because taking less than 5 drugs (n = 79), unable to provide an accurate list of the medications taken at home (n = 37), failed to provide an informed consent (n = 17) or had a life expectancy of less than 6 months (n = 21). Table 1 shows the main patient characteristics. Admissions were mainly from the emergency room (75% vs 15% from home) and the primary diagnoses that led to acute hospital admissions were mainly respiratory (34%) and cardiovascular diseases (31%), with a lower prevalence of neurological (13%), gastrointestinal (6%) and muscolo-skeletal diseases (6%). The goal of enrolling multimorbid elderly patients on polypharmacy was met, as witnessed by their mean age of 82±6.8 years and mean daily intake of 9.4 ± 3.7 drugs/patient. The time of patient permanence in the wards ranged from 9 to 12 days, being no longer than that recorded in the same wards before starting the study.
Table 1Main characteristics at admission of the 90 patients included in the study.
Females, No. (%)
Age (years), mean (±SD)
82 ± 6.8
Age ≥ 85 years, No. (%)
Education (years), mean (±SD)
9.4 ± 4.2
Widow/er, No. (%)
Living alone, No. (%)
Length of hospital stay (days)(mean (+SD)
Number of comorbidities, mean (±SD)
CIRS-Comorbidity index at admission, mean (SD)
3.3 ± 2.1
CIRS-Severity index at admission, mean (SD)
1.7 ± 0.3
Hypertension (ICD-9 codes: 401–405)
Coronary heart disease (ICD-9 codes: 410–414)
Heart failure (ICD-9 codes: 428)
Diabetes mellitus (ICD-9 codes: 250)
COPD/bronchial asthma (ICD-9 codes: 490–496)
Cerebrovascular disease (ICD-9 codes: 430–438)
Liver disease (ICD-9 codes: 570–579)
Osteoporosis (ICD-9 codes: 733) and Osteoarthritis (ICD-9 codes: 715)
Chronic renal failure (ICD-9 codes: 585)
Malignancy (ICD-9 codes: 140–239)
Medications at hospital admission
Number of drugs, mean (+SD)
9.4 ± 3.7
10 most frequently prescribed drugs (IV level of ATC classification):
Among a total number of 911 drugs prescribed to the 90 patients at the time of ward admission by the attending clinicians, who took into account those taken at home and in the emergency room, there were 259 discrepancies identified by the pharmacists and communicated to the ward clinicians, with a rate of 28% for all the prescriptions and a mean number of 3 discrepancies per patient. Table 2 shows that the majority of the discrepancies at admission were intentional and documented (68%), those intentional but not documented being 27% and the unintentional ones 5%. Table 1 also shows that the most frequent actions jointly taken by pharmacists and clinicians were drug deprescribing, particularly frequent as an intentional and documented action, as well as changes with another active principle of the same pharmacological class.
Table 2Number and types of medication discrepancies between pharmacists and clinicians at the time of hospital admission and the resulting actions.
Type of discrepancy
Intentional, not documented
Type of action
- Drug deprescribed
- Change with a drug of the same pharmacological class
- Change with a drug of a different but similar pharmacological class
Table 3 shows the number and types of DRPs identified during recognition/reconciliation. The most frequent, with roughly the same rate of detection, were prescription of an inappropriate drug, presence of DDIs of the highest D level and those of C level of major clinical relevance, whereas other types of DRPs were less frequent (risk of adverse reactions, over-dosage, duplicates or off-label use). The 10 drugs mainly involved in inappropriate use were: ticlopidine (6% of cases), antiarrhythmic drugs (6%), alpha1 blockers (3%), central alpha blockers (3%), benzodiazepines (3%), sulfonylureas (2%), antipsychotics (2%), barbiturates (1%), non-benzodiazepine hypnotics (1%) and tricyclic antidepressants (1%). The drug combinations most frequently involved in potentially severe DDIs (D and C levels) were: potassium-sparing diuretics and ACE-inhibitors (increased risk of hyperkalemia), 15% of cases; aspirin (low dose) and clopidogrel or ticlopidine (increased risk of bleeding), 10%; statins and calcium antagonist (increased risk of myopathy including rhabdomyolysis), 9%; amiodarone and beta-blocker (hypotension, bradycardia, or cardiac arrest), 7%; digoxin and spironolactone (increased risk of digoxin toxicity), 5%; clopidogrel and proton pump inhibitor (reduction in clinical efficacy of clopidogrel and increased risk for thrombosis), 5%; allopurinol and enalapril (hypersensitivity reactions), 5%; simvastatin and amiodarone (increased risk of rhabdomyolysis), 2%; digoxin and hydrochlorothiazide (increased risk of digoxin toxicity), 2%; potassium and potassium-sparing diuretics (increased risk of hyperkalemia), 1%.
Table 3Number and types of drug related problems (DRPs) identified in the frame of the recognition/reconciliation process at hospital discharge (n = 455) and consequences of the recognition/reconciliation process involving pharmacists and clinicians in relation to the type of drug-related problems (DRPs).
Pertaining to the acceptance by the ward clinicians of the recognition/reconciliation process, Table 2 shows that they accepted the changes suggested by the pharmacists in approximately two thirds of the instances, particularly when the identified DRPs were the prescription of contraindicated medications that led to deprescribing and when clinically dangerous DDIs were detected. If the pharmacists’ advice was accepted, the most frequent ensuing action was deprescribing the implicated drugs or at least their closer monitoring, particularly when the DRPs communicated by the pharmacists to clinicians was the use of drugs inappropriate for older people according to the STOPP/BEERS criteria, or when the risk of DDIs was considered major (D and C levels) according to INTERCheck.
The acquisition of expertise in the management of polypharmacy in older multimorbid medical patients, more and more frequent in the community as well as in the hospital settings, is an important goal in order to improve medication safety and prevent or at least reduce such negative outcomes as adverse drug events, that in turn lead to more frequent falls, re-hospitalizations and a higher death rate [
]. Thus, the present study was designed and carried out to assess whether or not it was feasible and acceptable for hospital clinicians to enlist for this purpose the driving role of hospital pharmacists, owing to their expertise in medication dynamics and kinetics, in the context of a structured process of drug recognition and reconciliation carried out through the close integration and collaboration of their different clinical expertises [
]. The underlying rationale was that the joint recognition/reconciliation process should help to identify and handle poor drug prescription appropriateness at the critical times of hospital admission and discharge. The main criteria used by the pharmacists in the frame of the reconciliation process were dosage appropriateness, the undue presence of drug duplicates and, most importantly, the use in older people of potentially inappropriate drugs according to the STOPP and BEERS criteria. Because both pharmacists and clinicians may have limited expertise in clinical pharmacology to implement these criteria in the reconciliation process, the practical advantage for those involved in this study was that useful information could also be obtained by the computerized support system INTERCheck. This tool was previously validated in the clinical context [
], with the demonstration that when the changes suggested by system were implemented by the attending clinicians the new onset and the potentially severe consequences of inappropriate drug prescriptions did decrease.
The results obtained in the present study should be discussed for three different aspects: the main and most frequent type of drug-related problems identified by the pharmacist with the aid of the information technology tool, the degree of acceptance of pharmacists’ suggestions by the clinicians and the actions jointly taken upon the identified drug-related problems.
Pertaining to the rate and types of drug-related problems the most frequent were the use of medications definitely inappropriate in older patients according to the well established criteria of STOPP and BEERS, as well as the detection of drug-drug interactions so potentially harmful in terms of clinical consequences to demand contraindication of the combination at stake. In agreement with the potential clinical relevance of the identified DRPs [
], at least two thirds of the related suggestions made by the pharmacists were accepted by the clinicians. The suggestions were also accepted with a high rate of agreement when drug overdosage and the presence of duplicates were identified. In the less frequent instances of non-acceptance of the pharmacists’ suggestions, the clinicians explained to them why they judged necessary to continue the medications actually involved in drug-drug interactions, but stated that they were prompted to implement a higher degree of monitoring of both the expected benefits and the potential adverse effects.
Pertaining to the issue of acceptance of the offered advice, in general clinicians judged useful the help provided by the hospital pharmacists with the help of the computer-based tool. As expected, there were a number of discrepancies between the prescriptions made by clinicians and changes suggested by the pharmacists, but the smooth interactions between these professionals, as well as the use of INTERCheck helped to reconcile the clinicians with the suggestions made by the pharmacists. Furthermore, the clinicians did not feel diminished in their professional capacity by the intervention of the pharmacists: not only because they did appreciate their contributes but also because they knew to be the ultimate decision maker, with the option to give up the suggestions on the basis of their clinical judgment and deeper knowledge of the patient status.
In conclusion, we established the feasibility of a hospital trajectory meant to promote a higher integration between pharmacists and clinicians attending acute medical patients, with the ultimate goal of optimizing drug prescription safety. This goal is consistent with the WHO Global Patient Safety Challenge [
WHO The Third WHO global patient safety challenge.
], that strongly encourages actions meant to reduce severe and avoidable medication harms. Strengths of this study are its prospective design and a sample size adequate to answer its main purpose, i.e., process feasibility. A limitation is the fact that it does not allow to establish whether or not the pharmacist-driven intervention is associated with benefits in terms of hard clinical outcomes such as the rate of adverse events, rehospitalization and mortality, because no follow-up was contemplated after hospital discharge in this pilot feasibility study.
Declaration of Competing Interest
PMM: Member of the scientific board for the Bayer Awards. He has also received from Bayer, Kedrion and Novo Nordisk honoraria for lectures at educational symposia. The other authors had no conflict of interest.
This work was partially supported by the Italian Ministry of Health- Bando Ricerca Corrente.
Appendix. Investigators and co-authors of the ad hoc Deprescribing Study Group are as follows
Cinzia Hu, Francesca Minonzio (Unità Operativa Medicina Interna IA); Anna Ludovica Fracanzani, Federica Iuculanno (Unità Operativa Medicine Interna IB); Nicola Montano, Eleonora Tobaldini (Unità Operativa di Medicina Generale Immunologia e Allergologia); Daniela Mari, Paolo Rossi, (UOC Geriatria, Fondazione IRCCS Ca’ Granda Ospedale Maggiore Policlinico); Valter Monzani, Francesca Tantardini, (Unità Operativa di Medicina d’Urgenza); Flora Peyvandi, Raffaella Rossio, (Unità Operativa di Medicina GeneraleEmostasi e Trombosi).
REPOSI Collaborators (REPOSI is the acronym of REgistro POliterapie SIMI, Società Italiana di Medicina Interna)
Polypharmacy in older people: lessons from 10 years of experience with the REPOSI register.