Article Text
PDF
Abstract
Objectives Older adults are a vulnerable and growing segment of the population with a high burden of comorbid conditions. As a consequence of increased co-morbidity, drug use in older adults is high, and polypharmacy has been linked to higher risk of adverse drug–drug interactions, morbidity, and mortality. The aim of the study was to evaluate the prevalence and nature of potentially inappropriate medications (PIMs) in a group of hospitalised older patients, and verify whether the use of ‘Beers criteria’ and/or ‘START & STOPP’ criteria could lead to deprescribing of drugs and reduce the length of stay in the hospital.
Methods Two hundred acutely ill patients aged ≥65 with multimorbidity admitted to the Division of Internal Medicine were enrolled in the study. Enrolled patients were admitted as medical emergencies and observed during their hospitalisation at the Emergency Department and subsequently at the Division of Internal Medicine. The pharmacological treatments taken by patients at home, during hospitalisation and at discharge, were examined, identifying inappropriate prescriptions (IPs), according to ‘Beers criteria’ and ‘START and STOPP’ criteria.
Results There were 487 IPs: 175 according to the Beers criteria; 50 according to the STOPP criteria; one according to the START criteria; 107 major interactions; 152 minor interactions; one off-label drug; and one duplicated pharmacotherapy. Twenty-three adverse drug reactions (ADRs) were recorded: the most frequent were abnormalities of serum electrolytes (35%); haemorrhagic events (22%); and accidental falls from benzodiazepine use (9%). The correct application of these criteria decreased IPs by 38% and reduced the number of drugs prescribed by the physician during the stay in the medical ward and at discharge by 19%.
Conclusions The use of criteria that detect IPs reduced PIMs and ADRs, increased safety in older patients, and reduced the number of drugs prescribed but did not reduce the length of stay in hospital.
- geriatric medicine
- quality in health care
- therapeutic drug monitoring
- medical errors
- competency evaluation
Data availability statement
Data are available upon reasonable request. Individual participant data that underlie the results reported in this manuscript will be available after deidentification.
Statistics from Altmetric.com
- geriatric medicine
- quality in health care
- therapeutic drug monitoring
- medical errors
- competency evaluation
Introduction
Pharmacological polytherapy has increased in recent decades due to the gradual ageing of the population and the greater availability of therapeutic options, especially in older frail patients (age ≥65 years).1 2 The current therapeutic approach, particularly for some diseases, is characterised by the use of multiple drugs in combination. The use of multiple drugs increases the likelihood of drug–drug interactions and adverse drug reactions (ADRs), as well as decreased adherence to therapy with a consequent reduction in efficacy.3 Interactions between drugs, and adverse events caused by errors in drug prescribing, could be responsible for increased morbidity, longer hospital stay, and an increase in hospital readmission or visits to the Emergency Department.4
Several sets of indicators have been developed in order to evaluate the quality of care in older adults, with the ultimate goal being to identify drugs to be avoided, potential drug interactions, and the prescribing cascade.5 The 2015 American Geriatrics Society Beers Criteria includes lists of potentially inappropriate medications to be avoided in older adults.6 It is important to highlight that the Beers List is not a substitute for professional judgement in prescribing decisions for the older adults, but it is intended to guide clinical management. The clinical judgement of the prescriber on an individual patient’s needs should always be considered.7 The Screening Tool of Older People’s potentially inappropriate Prescriptions (STOPP) is a European tool addressing overprescribing in older patients, whereas the Screening Tool to Alert doctors to Right Treatment (START) is an evidence-based screening tool to detect prescribing omissions in older adults.8 9 The ‘Beers criteria’ and ‘START’ and ‘STOPP’ tools are valid for the identification of inappropriately prescribed pharmacotherapy in older patients.
A transition from one healthcare setting to another (hospitalisation, discharge, or during intra-hospital transfer, for example from intensive care units to inpatient wards) increases the risk of medication errors, especially in older patients. ‘Care transitions’ may result in harm to the patient due to adverse drug events (as a result of inappropriate changes in therapy), increased duration of hospital stay, or early readmission after discharge. Several strategies have previously been applied to improve care transitions and reduce adverse clinical outcomes. Involving pharmacists during hospitalisation has been frequently studied, however the effect on outcomes is variable.10
The percentage of patients with at least one medication discrepancy varies considerably between studies: between 3.4% and 97.0% at admission and between 25% and 80% at discharge.11 The Italian Ministry of Health promoted Recommendation no. 17, “Recommendation for the reconciliation of drug therapy”, in December 2014, with the aim of preventing errors in therapy at care transitions (Ministero della Salute 2014).
The present study aims to determine, in acutely ill older/frail patients, the prevalence and nature of IPs by analysing pre-admission drug therapy, medications taken during hospital stay in the Emergency Department and in the Internal Medicine ward, and drugs prescribed at discharge. Additionally, another scope of the study was to verify if the procedure of pharmacological reconciliation led to a reduction in the number of drugs being prescribed to the patients and to a reduced length of hospital stay.
Patients and methods
Two hundred acutely ill older frail patients aged ≥65 with multimorbidity, consecutively admitted to the Division of Internal Medicine at Antonio Cardarelli Hospital (a 900-bed tertiary care hospital), Naples, Italy, from January to June 2018 were enrolled in the study. Enrolled patients were admitted as medical emergencies and observed during their hospitalisation at the Emergency Department and subsequently at the Division of Internal Medicine. Multimorbidity was defined as three or more different morbidities for which they were under care. An internist with experience in the management of geriatric patients (AB) assessed the patient’s multimorbidity using the Cumulative Illness Rating Scale for Geriatrics. We excluded subjects suffering from terminal neoplastic pathologies and patients with a length of hospital stay <3 days. In our internal medicine unit, a computerised prescription drug system is currently employed. The study was performed by two postgraduate pharmacists (MM and MRD) and a physician with expertise in geriatric care (AB). All investigators were trained in the use of the Beers criteria and START & STOPP tools, and were subsequently monitored by a senior clinical pharmacist (MLA) and the Director of the Internal Medicine Unit (AI). One hundred patients enrolled from January to March 2018 were consecutively admitted to the usual care cohort (control group). In April to June 2018, 100 patients were consecutively admitted to the intervention cohort, in which they were subjected to a systematic medication review. The Cumulative Illness Rating Scale was not different between the two groups. However, the intervention group was older than the control group (79.1±7.0 vs 76.7±7.0 years, [mean±SD], P<0.02). Informed consent from study participants was not considered necessary for this study. Therefore, an accurate review of medicines was performed, including those taken by the patient at home and during hospital stay for the new clinical condition.
The review of patient’s pharmacotherapy was organised in two phases: the first phase, medication recognition, consists of collecting a patient’s personal data, history of alcohol intake, smoking history, and known allergies, followed by acquiring an accurate list of medications, including a check for the appropriateness of prescriptions and documentation of any changes. As well as this, a list of any other substances taken by the patient, such as over-the-counter medications, homeopathic remedies, supplements, and herbal remedies was acquired.
The second phase, medication reconciliation, ensures that the drugs the patient should be prescribed match those that are prescribed. Before prescribing a drug, the data collected through the medicine recognition phase were analysed and reconciliation was performed. Box 1 shows the types of inappropriate prescriptions (IPs). The pharmacist monitoring medications collected data on the pharmacological therapies of all patients over 65, according to the following steps: clinical interview including a pharmacological survey within 24–48 hours of admission in the internal medicine unit; evaluation of any IPs (using the STOPP and Beers criteria); identification of drug or food interactions, using the DrugReax database of Micromedex, differentiating them into major and minor interactions; evaluation of drug discrepancies, both unintentional and intentional, together with the physician, in order to modify drug therapy both during hospitalisation and subsequent discharge (eventually applying the START criteria); pharmacological reconciliation and counselling the patient on how to correctly adhere to any new drug therapy that is continued after discharge from hospital; and detection of any ADRs and their inclusion in the National Pharmacovigilance Network.
Types of inappropriate prescriptions
Off-label use, intended as use not in accordance with the package label of the medicine (in terms of indication, dose, or route of administration).
Interactions between different drugs or between drugs and food or alcohol.
Interactions or summative effects of drugs belonging to the same therapeutic category (pharmacological duplications).
Potentially inappropriate drugs in the older patient or drug interactions with concomitant diseases.
Prescription cascade (drugs being prescribed to treat unrecognised ADRs).
Statistical analysis
The differences between the two groups were found using variance analysis in which age and gender of the participants were inserted as covariates. The proportions were examined using a Chi-squared test. A value of P<0.05 was considered to be statistically significant.
Results
In total 200 patients were enrolled, equally divided by gender and all of white ethnic origin – table 1 shows the characteristics of the study sample. All admissions were for medical reasons, with pathology of the cardiovascular system being most prevalent, followed by respiratory, neurological, and haematological disease (table 1). There were no significant differences between the two groups. The most common comorbidity was hypertension (92%), followed by diabetes mellitus (34%) and chronic obstructive pulmonary disease (16.7%). The number of drugs taken by patients before admission to hospital was on average 7.5±3.22 (mean±SD) per patient. Few patients took less than five drugs (27%), with the remaining 73% taking more than five (taking polypharmacy). During hospital stay in the Internal Medicine ward 8.0±2.4 (mean±SD) drugs were prescribed at admission to the patients in the control group and 7.5±2.8 to the patients in the intervention group before the procedure of reconciliation. After reconciliation and at discharge the number of drugs in the intervention group decreased to 6.1±3.1 per patient (P=0.002 comparing number of drugs after reconciliation vs before reconciliation in the intervention group). Patients in the intervention group after reconciliation took two units of medicines less than patients in the control group at discharge (P<0.001).
Characteristics of the study sample (n=200)
Table 2 shows interactions between different drugs and prescriptions of potentially inappropriate drugs in older patients before and after reconciliation: major interactions between drugs reduced by 46% and the prescription of potentially inappropriate drugs in older patients showed a reduction of 38%. In particular, before reconciliation, there was one use of an off-label indication (use of Dabigatran in valvular atrial fibrillation), one case of duplicated pharmacotherapy (simultaneous use of three opioid drugs: tramadol, tapentadol and oxicodone), and one non-prescription of anticoagulant drugs in a patient with atrial fibrillation. The drugs most frequently used inappropriately before admission to the hospital were proton pump inhibitors (PPIs), included in the Beers list of criteria (drugs or pharmacological classes potentially harmful to the majority of the older adults regardless of their clinical condition): 58% used PPIs for more than 8 weeks. Acetylsalicylic acid was used by 29% of patients and furosemide by 25%, even though these medications are included in List c of the Beers Criteria (drugs to be used only with extreme caution). Online Supplementary table shows the drugs most frequently involved in the formation of major interactions: antiplatelet agents, antidepressants, diuretics, ACE inhibitors, and digoxin.
Supplemental material
Interactions between different drugs and IPs before and after reconciliation
A total of 23 potential ADRs were recorded (table 3). Most of them were classified as possibly or probably related to the drugs (Naranjo Algorithm). The most frequently found ADRs were abnormalities of serum electrolytes (n=8), haemorrhagic events (n=5, of which one was fatal), accidental falls from benzodiazepine use (n=2), and heart failure (n=2).
Adverse drug reaction
Discussion
Older adults are a vulnerable and growing segment of the population with a high burden of comorbid conditions. As a consequence of increased co-morbidity, drug use in older adults is high, and polypharmacy has been linked to a higher risk of adverse drug–drug interactions, morbidity, and mortality.11–14 Optimising the treatment of older patients is a complex challenge. The International Group for Reducing Inappropriate Medication Use & Polypharmacy recommends that a physician should coordinate decisions within a shared framework, preferably in a setting where multiple morbidity and polypharmacy can be adequately considered together with a pharmacist.15 Reviews that have assessed the effectiveness of medication reconciliation interventions in the hospital setting have found that it results in a reduction in medication discrepancies and adverse drug events, but reported no positive results on length of stay or post-discharge clinic visits.16 17 Key aspects of successful interventions include intensive pharmacy staff involvement and targeting intervention to a high-risk patient population, such as older adults.18 In the best hospital setting an interdisciplinary geriatric consultation team (IGCT), with physicians of various medical specialties, nurses, dieticians, and pharmacists, is regularly provided to ensure comprehensive geriatric assessments in older inpatients. A Belgian study demonstrated that a structured medication review as part of usual IGCT care may contribute to an increased detection of drug-related problems and help to further reduce polypharmacy in older inpatients.19 Coleman et al demonstrated that by improving transitions of care through visits with a care team that included physicians, nurses, and pharmacists the number of Emergency Department visits was reduced.20 21 A recent study in an internal medicine ward found that medication discrepancies were frequent both on admission and at discharge.22
In this article we describe the experience of a collaborative team of pharmacists and physicians in a ward of internal medicine in an Italian hospital. Hospital pharmacists facilitate and ensure the execution of many services, such as medication recognition and reconciliation, patient education, and medication review in collaboration with the physician. The information provided to the physicians in this study can represent a precious resource, contributing to improve the quality of drug prescription and to reduce IPs and possible adverse reactions. The attending physician examined any medications where the appropriateness of a prescription was queried. Reconciliation errors were defined as discrepancies considered by the physician to be unjustified. When such errors were found, the appropriate changes were made in the patient’s prescription. However, medication regimens at hospital discharge often differed from preadmission medications and some differences reflected deliberate changes related to the conditions that led to hospitalisation. However, other discrepancies were unintentional and resulted from incomplete or inaccurate information about current medications and doses. Fortunately, clinically significant unintentional discrepancies affect few patients.23 The aim of reconciliation is to ensure that patients correctly receive all drugs that they were taking before hospitalisation, monitoring that they were prescribed with the right dosage regimen (dosage, route, and frequency) and that pharmacological therapy is adapted to the current clinical condition of patients in hospital and to any new pharmacotherapy prescribed after the transition of care.24 25
This study has some limits. First, a relatively small number of patients were enrolled, and second the lack of inclusion in the reconciliation stage of drugs used in the emergency setting, which could be responsible for potential ADRs. However, it should be considered that, generally, these drugs are responsible for idiosyncratic type B reactions (eg, allergies) that, unless they were previously known by the patient, are not predictable in clinical practice. Finally, this is a descriptive study, so no statements can be made on effects of medication review on clinical or economical outcomes.
In conclusion the present study shows that criteria such as the Beers Criteria list and STOPP/START criteria should always be consulted when administering therapy to older patients. The correct application of these criteria in a clinical setting improved medication selection, educated clinicians and patients, reduced the number of drugs prescribed by the physician during the stay in the medical ward and at discharge, and served as a tool for judging quality of care and cost in the management of hospitalised older adults. However, in the present study the application of Beers Criteria and START & STOPP tools did not result in a diminished length of hospitalised stay. Ultimately, our data demonstrated that regardless of the criteria used, the prevalence of IPs used in the studied sample was high. These findings support the choice of Beers Criteria or STOPP as a medication review tool to improve care for older adults. Furthermore, close collaboration between pharmacists and physicians was found to be beneficial.
What this paper adds
What is already known on this subject
The interactions between drugs and adverse events caused by errors in drug prescribing could be responsible for increased morbidity, length of hospital admission, and an increase in hospital readmissions/visits to the Emergency Department. A transition from one healthcare setting to another (hospitalisation, discharge, or even when transferred from one department to another) increases the risk of medication errors, especially in older patients.
What this study adds
The correct application of Beers and START & STOPP criteria in a clinical setting improved medication selection, educated clinicians and patients, decreased adverse drug events, reduced the number of drugs prescribed by the physician during the stay in the medical ward and at discharge, even though it did not result in a diminished length of hospitalised stay.
Data availability statement
Data are available upon reasonable request. Individual participant data that underlie the results reported in this manuscript will be available after deidentification.
Acknowledgments
We thank Dr Giuseppe Chiariello, Dr Antonella Schettini, Dr Giuseppe Covetti, Dr Raimondo Cavallaro, Dr Maria Carla Pisano, Dr Tiziana D’Aniello, and Dr Rossella Nappo for their help in the realisation of this project.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
EAHP Statement 4: Clinical Pharmacy Services.
Contributors Study concept and design: MLA, AB, AI. Collection, management, analysis, and interpretation of the data: MLA, AB, MM, MRD, AI. Drafted or critically revised the manuscript for important intellectual content: MLA, AB, GG, ET, RM, AE, AI. All authors read and approved the final manuscript.
Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.
Competing interests None declared.
Provenance and peer review Not commissioned; externally peer reviewed.