Abstract
Objective
To determine the utility of a tool (TRIGGER-CHRON) for identifying adverse drug events (ADEs) associated with the administration of high-alert medications in elderly patients with multimorbidity and to determine the medications most frequently implicated.
Methods
A retrospective observational study was conducted at 12 Spanish hospitals. A random sample of five medical records from each hospital was selected weekly for review over a 12-week period. We included patients aged 65 and over with multimorbidities, hospitalised for >48 hours. ADEs detected by the 32 TRIGGER-CHRON signals and caused by high-alert medications included on the Spanish HAMC list for chronic patients were selected for analysis. Triggers identified and ADEs detected were recorded. The severity and preventability of the ADEs were evaluated. The positive predictive value (PPV) of each trigger was calculated.
Results
On 720 charts reviewed, 908 positive triggers were identified that led to the detection of 158 ADEs caused by at least one high-alert medication on the HAMC list. These ADEs occurred in 139 patients (prevalence 19.3/100 admissions). The majority of ADEs were mild and 59.5% were deemed preventable. The drugs most frequently associated with ADEs were corticosteroids, loop diuretics, opioid analgesics and oral anticoagulants. Fifteen triggers had PPVs ≥20%. Six triggers (serum glucose >110 mg/dL, abrupt cessation of medication, oversedation/lethargy, hypotension, adverse reaction recorded and constipation) accounted for 69.8% of the ADEs identified.
Conclusions
Applying the TRIGGER-CHRON to hospitalised patients with multimorbidity in 12 Spanish centres allowed detection of one adverse event caused by a high-alert drug for every four patients, which were preventable in a large proportion of patients. This confirms the need to establish interventions that reduce harm with these medications. We believe that TRIGGER-CHRON can be a useful tool to measure this harm and to determine the effects of medication safety improvement programmes as they are implemented.
Keywords: high-alert medications, trigger tool, drug-related side effects and adverse reactions/diagnosis, multimorbidity, patient safety
Introduction
Adverse drug events (ADEs) have been highlighted as a major patient safety and public health challenge.1 Most of these events are linked to unsafe medication practices and medication errors and, consequently, are considered preventable. Recognising this problem, the WHO launched its third Global Patient Safety Challenge: Medication Without Harm in March 2017, with the goal of reducing severe avoidable medication-related harm by 50% over the next 5 years, globally.2 The WHO has asked countries and key stakeholders to prioritise three areas that are associated with high medication error rates for early actions: transitions of care, inappropriate polypharmacy and high-risk situations, which includes the use of high-alert medications.3
The concept of 'high-alert or high-risk medications', referring to those medications that bear a heightened risk of causing significant patient harm when used in error, is a key concept in patient safety. The Institute for Safe Medication Practices (ISMP) introduced this concept in 1998 after conducting a study which revealed that a relatively small number of medications was responsible for the majority of medication errors resulting in serious consequences for patients.4 Hence, it was proposed that efforts should focus on improving the safe use of these medications to make them more efficient and to avoid preventable harm to patients.5 Specific medications considered high-alert drugs may differ because the medications most likely associated with harm vary depending on the setting and disease epidemiology.6
Polypharmacy is a serious health problem that continues to gain importance because of ageing of populations and, consequently, the increase in chronic patients with multimorbidity who need to take multiple medications.7 8 Several studies have shown that the risk of medication errors increases with the number of medications prescribed due to drug interactions, suboptimal patient adherence and reduced quality of life.9–11 The risk of these errors leading to adverse events, sometimes severe, will be greater if high-alert medications are involved. Therefore, the WHO believes that the priority areas of action are not mutually exclusive3; healthcare professionals should concentrate their efforts particularly on the most vulnerable patients, such as multimorbidity patients who have been prescribed high-alert medications and who are at a higher risk of serious harm from medications.
A key component for achieving the objective of this third WHO challenge is measuring the impact of the interventions performed to track progress towards improvement, which requires an efficient tool to detect ADEs in the specific areas on which this challenge focuses. TRIGGER-CHRON has recently been validated as a tool that consists of several triggers that provide an easy way to identify ADEs in elderly patients with multimorbidity.11 A trigger is a flag or prompt that promotes a focused and selective process for screening a patient’s medical record review to determine the presence or absence of an ADE. For instance, the 'administration of naloxone' is a trigger that may indicate an overdose of an opioid with respiratory depression (an ADE). Triggers can be used to obtain information about ADE rates over time at institutions and to monitor the impact of interventions.12 On the other hand, in Spain there is a list of high-risk medications specifically for chronic patients, the HAMC list, which was developed by the Spanish National Health System.13
The main objective of this study was to determine the utility of TRIGGER-CHRON for identifying ADEs associated with the administration of high-alert medications on the HAMC list for elderly chronic patients with multimorbidity and for determining the medications commonly responsible for ADEs.
Methods
The study was performed using clinical data obtained from a previous multicentre study carried out to validate the TRIGGER-CHRON tool in elderly patients with multimorbidity.11 This was a retrospective observational study that was conducted in 12 Spanish hospitals over a 12-week period (20 March to 11 June 2017). At each hospital a weekly sample of five medical records of patients discharged the week before from internal medicine or geriatric units was selected for review using the randomisation function found at https://www.random.org/sequences/. Patients were eligible if they were 65 or older, suffered from multimorbiditiy (defined as the presence of two or more chronic medical conditions)14 and had a hospital stay >48 hours. Patients were excluded if they were hospitalised in clinical units other than internal medicine or geriatrics, were receiving palliative care, or had been transferred from other clinical units and not from any of the units under study (eg, ICU).
In brief, the procedure was as follows. First, the selected charts were reviewed in each hospital by a clinical pharmacist for the presence of triggers. Second, when a trigger was found, the patient record was closely analysed to determine whether or not an associated ADE had occurred. An ADE was defined as any injury resulting from medical interventions related to a drug; this includes both adverse drug reactions in which no error occurred and complications resulting from medication errors.15 Third, the detected ADEs that occurred in patients during their hospitalisation were assessed to determine their severity and preventability. Those ADEs that had contributed to or caused hospitalisation were eliminated from the study. The severity of the ADEs was evaluated using the National Coordinating Council for Medication Error Reporting and Prevention Index (NCCMERP).16 The TRIGGER-CHRON counts only ADEs—that is, harm to the patient from medications. Accordingly, only categories E to I were used because these categories imply harm: E (temporary harm to the patient that required intervention), F (temporary harm to the patient that required or prolonged hospitalisation), G (permanent patient harm), H (intervention required to sustain life) and I (patient death). ADE preventability was evaluated using the Schumock and Thornton criteria adapted by our working group.17 18 Whenever questions or discrepancies arose, the hospital physician responsible for the patient involved was contacted. Finally, the following variables from each record were recorded using an electronic data collection tool developed for the study: age, sex, number and types of chronic diseases, length of hospital stay, medications taken at home and administered during the hospital stay, triggers identified and characteristics of the ADEs detected (triggers associated, medications involved, severity and preventability). All data were reviewed by the principal investigators for consistency. All questions were directed to the responsible investigator at each hospital for resolution.
The study protocol was authorised by the Ethics Committee of Clinical Research at the University Hospital Virgen del Rocío located in Seville, Spain (coordinating hospital) and then again by the local ethics committees at each participating hospital, according to Spanish regulations.
In the validation study, a set of 51 triggers was used to review the medical records, of which only 32 triggers were finally selected to integrate into the TRIGGER-CHRON since it was found that the rest of the triggers were not efficient for detecting ADEs.11 To carry out the present study, aimed at evaluating the usefulness of TRIGGER-CHRON to detect ADEs associated with high-alert medications, only ADEs that were detected by the 32 TRIGGER-CHRON signals and that were caused by the high-alert medications included on the HAMC list (box 1) were selected for analysis with one exception: all the ADEs detected that were caused by the use of systemic corticosteroids were selected rather than just corticosteroids after long-term use, as indicated by the HAMC list, as this was a study carried out in a shorter period of time. Table 1 shows the 32 triggers included in the TRIGGER-CHRON organised into five modules: 7 care module triggers, 7 antidotes/drug treatments, 2 medication concentrations, 15 abnormal laboratory values, and 1 emergency department trigger. The table shows the high-alert medications whose most frequent ADEs could be detected using some of the specific TRIGGER-CHRON triggers. The tool also includes five general triggers that allow for identifying ADEs due to any medication (C1, C2, D2, D7 and E1).
Box 1. List of high-alert medications for patients with chronic diseases (HAMC list).13 .
Therapeutic clases
Anticoagulants, oral
Antiepileptics (narrow therapeutic range)
Antiplatelets (including aspirin)
Antipsychotics
β-Adrenergic blockers
Benzodiazepines and analogues
Corticosteroids long-term use
Cytostatic drugs, oral
Immunosuppressants
Insulins
Loop diuretics
Non-steroidal anti-inflammatory drugs
Oral hypoglycaemic drugs
Opioid analgesics
Specific medications
Amiodarone/ dronedarone
Digoxin oral
Methotrexate, oral (non-oncologic use)
Spironolactone/eplerenone
Table 1.
List of triggers included in the TRIGGER-CHRON and high-alert medications whose potential adverse drug events could be detected using these triggers
| TRIGGER-CHRON | High-alert medications | |
| Module 1: Care triggers | ||
| C1 | Rash | All medications |
| C2 | New allergy | All medications |
| C3 | Oversedation/lethargy | Antipsychotics, benzodiazepines, opioids |
| C4 | Hypotension | β-adrenergic blockers, loop diuretics, spironolactone/eplerenone |
| C5 | Transfusion or use of blood products | Oral anticoagulants, antiplatelets, NSAIDs |
| C6 | Constipation | Opioids |
| C7 | Adverse reaction recorded | All medications |
| Module 2: Antidotes/drug treatments | ||
| D1 | Vitamin K administration | Oral anticoagulants |
| D2 | Antihistamines IV | All medications |
| D3 | Flumazenil administration | Benzodiazepines |
| D4 | Naloxone administration | Opioids |
| D5 | Antiemetic administration | Cytostatic drugs, opioids, others |
| D6 | Haloperidol administration | Opioids |
| D7 | Abrupt cessation of medication | All medications |
| Module 3: Medication plasma concentration triggers | ||
| P1 | Digoxin level >2 ng/mL | Digoxin |
| P2 | Carbamazepine >12 µg/mL | Antiepileptics |
| Module 4: Laboratory results triggers | ||
| L1 | Clostridium difficile positive stool | – |
| L2 | Serum glucose <50 mg/dL | Insulins, oral hypoglycaemic drugs |
| L3 | Serum glucose >110 mg/dL | Corticosteroids, insulins |
| L4 | INR >5 | Oral anticoagulants |
| L5 | Rising BUN or serum creatinine >2 times baseline | Loop diuretics, NSAIDs, amiodarone/dronedarone, spironolactone/eplerenone |
| L6 | eGFR <35 mL/min/1.73m2 | Loop diuretics, immunosuppressants, NSAIDs, spironolactone/eplerenone |
| L7 | K >6.0 mEq/L | Immunosuppressants, loop diuretics, spironolactone/eplerenone |
| L8 | K <2.9 mEq/L | Loop diuretics |
| L9 | Na <130 mEq/L | Loop diuretics, spironolactone/eplerenone |
| L10 | ALT >80 U/L and AST >84 U/L | Cytostatic drugs, corticosteroids |
| L11 | ALP >350 U/L and total bilirubin >4 mg/dL | Cytostatic drugs, corticosteroids |
| L12 | CPK >269 U/L | Cytostatic drugs |
| L13 | TSH <0.34 μUI/L or T4 >12 µg/dL | Amiodarone/dronedarone, cytostatic drugs |
| L14 | HbA1c >6% and glucocorticoid | Corticosteroids |
| L15 | White blood cells <3000 | Cytostatic drugs, immunosuppressants, methotrexate |
| Module 5. Emergency department (ED) triggers | ||
| E1 | Readmission to ED within 48 hours | All medications |
ALP, alkaline phosphatase; ALT, alanine aminotransferase; BUN, blood urea nitrogen; CPK, creatine phosphokinase; GFR, glomerular filtration rate; HbA1c, glycated haemoglobin; INR, International normalised ratio; NSAIDs, non-steroidal anti-inflammatory drugs; TSH, thyroid-stimulating hormone.
The positive predictive value (PPV) was calculated for each trigger as the number of ADEs identified using this trigger divided by the number of times the trigger was identified in the records.
Descriptive statistics were calculated for patients and ADE characteristics. Categorical variables were summarised as percentages and continuous variables as median (range).
Results
In the 720 medical records that were reviewed (60 charts per hospital), a total of 1247 positive triggers of the TRIGGER-CHRON were found, which allowed for the detection of 212 ADEs. Of these, 158 (74.5%) ADEs were caused by at least one high-alert medication on the HAMC list and occurred in 139 patients, which implies a prevalence of adverse events associated with high-alert medications of 19.3 per 100 admissions. The median age of the patients was 84 years (range 65–99), of whom 58.2% were women. The median number of chronic illnesses per patient was 6 (range 2–12), the median number of medications taken by patients before admission was 9 (range 3–19) and during their hospital stay was 17 (range 7–37), and their mean±SD hospital stay was 10.2±7.1 days.
In total there were 167 high-alert medications associated with the 158 ADEs (table 2). The most commonly involved drugs were corticosteroids (38 ADEs, 24.1%), loop diuretics (30 ADEs, 19%), opioid analgesics (26 ADEs, 16.5%) and oral anticoagulants (21 ADEs, 13.3%). The majority of ADEs were categorised as NCCMERP harm category E (n=144); 21 ADEs were in category F, whereas only 1 ADE was classified as G and another as H. A total of 94 ADEs were potentially preventable (59.5%; 94/158). The medications most frequently associated with preventable ADEs were also corticosteroids, opioids, anticoagulants and loop diuretics.
Table 2.
High-alert medications involved, severity and preventability of adverse drug events (ADEs) detected
| High-alert medications involved | ADEs detected* n=158 (%) |
Severity score† | Preventability of ADEs | ||||
| E | F | G | H | Yes, n=94 (59.5%) | No, n=64 (40.5%) | ||
| Corticosteroids | 38 (24.1%) | 37 | 1 | – | – | 21 | 17 |
| Loop diuretics | 30 (19.0%) | 27 | 3 | – | – | 15 | 15 |
| Opioid analgesics | 26 (16.5%) | 23 | 2 | 1 | – | 21 | 5 |
| Anticoagulants oral | 21 (13.3%) | 15 | 6 | – | – | 16 | 5 |
| Antipsychotics | 14 (8.9%) | 13 | 1 | – | – | 9 | 5 |
| Spironolactone/eplerenone | 9 (5.7%) | 9 | – | – | – | 3 | 6 |
| Antiplatelets (including aspirin) | 7 (4.4%) | 3 | 4 | – | – | 2 | 5 |
| Benzodiazepines and analogues | 7 (4.4%) | 7 | – | – | – | 7 | -- |
| Insulins | 5 (3.2%) | 3 | 1 | – | 1 | 3 | 2 |
| β-Adrenergic blockers | 3 (1.9%) | 3 | – | – | – | -- | 3 |
| Oral hypoglycaemic drugs | 2 (1.3%) | – | 2 | – | – | 1 | 1 |
| Digoxin | 2 (1.3%) | 2 | – | – | – | 2 | -- |
| Antiepileptics (narrow therapeutic range) | 1 (0.6%) | 1 | – | – | – | 1 | -- |
| Non-steroidal anti-inflammatory drugs | 1 (0.6%) | 1 | – | – | – | – | 1 |
| Immunosuppressants | 1 (0.6%) | – | 1 | – | – | 1 | -- |
| Total | 167 | 144 | 21 | 1 | 1 | 102 | 65 |
*In some cases, more than one drug is involved in one ADE.
†NCCMERP level of harm. E: temporary harm to the patient and required intervention; F: temporary harm to the patient and required initial or prolonged hospitalization; G: permanent patient harm; H:intervention necessary to sustain life.
Table 3 shows, for each trigger, the number of ADEs that allowed identification, the number of times the trigger was found after reviewing the medical records and its PPV, and the high-alert medications involved in the ADEs detected through the trigger. Some ADEs were identified by more than one trigger. For example, ADEs caused by oral anticoagulants were identified through the specific triggers 'INR >5', 'vitamin K administration' and 'transfusion or use of blood products' and also by the general triggers 'abrupt cessation of medication' and 'adverse reaction recorded'. A total of 25 of the 32 TRIGGER-CHRON triggers were found in the cards, although individual triggers varied in their yield for ADE identification. Fifteen triggers had an efficiency, evaluated in terms of the PPV, equal to or greater than 20%. Triggers that allowed for identification of a greater number of ADEs were the specific triggers 'serum glucose >110 mg/dL' (n=38), 'oversedation/lethargy' (n=22), 'hypotension' (n=19) and 'constipation' (n=18) with PPVs of 20.9%, 30.9%, 33.9% and 25%, respectively. Two general triggers—'abrupt cessation of medication' and 'adverse reaction recorded'—also allowed for detection of a high number of ADEs and had PPVs ≥50%. These six triggers were found on 69.8% (148/212) of the ADEs that were identified.
Table 3.
Prevalence and positive predictive value (PPV) of triggers and medications involved in the adverse drug events (ADEs) detected
| Trigger | No of ADEs identified with the trigger* | No of triggers found in the charts | PPV (%) | High-alert medications involved in the ADEs detected (n)† | |
| L3 | Serum glucose >110 mg/dL | 38 | 182 | 20.9 | Corticosteroids (36), insulins (2) |
| D7 | Abrupt cessation of medication | 32 | 60 | 53.3 | OAC (8), loop diuretics (7), antiplatelets (5), antipsychotics (4), benzodiazepines (3), β- blockers (2), spironolactone (2), insulins (1) |
| C3 | Oversedation/lethargy | 22 | 71 | 30.9 | Antipsychotics (13), benzodiazepines (5), opioids (4) |
| C4 | Hypotension | 19 | 56 | 33.9 | Loop diuretics (15), β-blockers (3), spironolactone (1) |
| C7 | Adverse reaction recorded | 19 | 38 | 50 | OAC (5), loop diuretics (3), opioids (3), benzodiazepines (2), oral hypoglycaemic (2), spironolactone (2), antiplatelets (1), NSAIDs (1) |
| C6 | Constipation | 18 | 72 | 25 | Opioids (18) |
| L4 | INR >5 | 8 | 25 | 32 | OAC (8) |
| L6 | eGFR <35 mL/min/1.73 m2 | 7 | 99 | 7.1 | Loop diuretics (4), spironolactone (2), immunosuppressants (1) |
| D1 | Vitamin K administration | 6 | 30 | 20 | OAC (6) |
| L8 | K <2.9 mEq/L | 6 | 20 | 30 | Loop diuretics (6) |
| C5 | Transfusion or use of blood products | 5 | 63 | 7.9 | Antiplatelets (3), OAC (2) |
| L7 | K >6.0 mEq/L | 5 | 18 | 27.7 | Loop diuretics (3), spironolactone (2) |
| L14 | HbA1c >6% and glucocorticoid | 5 | 7 | 71.4 | Corticosteroids (5) |
| D5 | Antiemetic administration | 4 | 31 | 12.9 | Opioids (3), digoxin (1) |
| D3 | Flumazenil administration | 3 | 5 | 60 | Benzodiazepines (3) |
| D6 | Haloperidol administration | 3 | 50 | 6 | Antipsychotics (2), opioids (1) |
| L2 | Serum glucose <50 mg/dL | 2 | 16 | 12.5 | Insulins (2) |
| D4 | Naloxone administration | 2 | 4 | 50 | Opioids (2) |
| P1 | Digoxin level >2 ng/mL | 2 | 8 | 25 | Digoxin (2) |
| P2 | Carbamazepine >12 µg/mL | 1 | 1 | 100 | Antiepileptics (1) |
| L5 | Rising BUN or serum creatinine >2 times baseline | 1 | 11 | 9.1 | Loop diuretics (1) |
| L9 | Na <130 mEq/L | 1 | 8 | 12.5 | Spironolactone (1) |
| L10 | ALT >80 U/L and AST >84 U/L | 1 | 8 | 12.5 | Corticosteroids (1) |
| L12 | CPK >269 U/L | 1 | 16 | 6.3 | Corticosteroids (1) |
| E1 | Readmission to ED within 48 hours | 1 | 9 | 11.1 | Loop diuretics (1) |
| Total | 212 | 908 | |||
*In several cases one ADE was identified with more than one trigger.
†In several cases more than one drug was involved in one ADE.
ALP, alkaline phosphatase; ALT, alanine aminotransferase; BUN, blood urea nitrogen; CPK, creatine phosphokinase; GFR, glomerular filtration rate; HbA1c, glycated haemoglobin; INR, International normalised ratio; NSAIDs, non-steroidal anti-inflammatory drugs; OAC, oral anticoagulants.
Discussion
ADEs are a common cause of morbidity and mortality worldwide. According to WHO priorities, in order to reduce preventable ADEs, actions should focus on three areas: transitions of care, polymedicated patients and high-risk situations. These areas are associated with a high risk of errors that cause harm to patients, so more efficient actions in these areas may be a way to actually reduce preventable harm from medications.3 In the present study, by using the TRIGGER-CHRON retrospectively in a random sample of elderly patients with multimorbidity, clinical pharmacists found that a high percentage of these patients (19.3%) suffered harm caused by high-alert medications of which a high proportion were preventable. Furthermore, the ADEs associated with high-risk medications comprised 74.5% of the total events attributed to all types of medications recorded. These findings underscore the importance of prioritising actions for this group of medications.
Our results also seem to indicate that TRIGGER-CHRON is an appropriate tool for capturing ADEs caused by high-alert medications in hospitalised patients with multimorbidity and for investigating the effect of interventions designed to decrease the number of incidents caused by these medications. It is important to point out that these interventions necessitate the combined knowledge and involvement of pharmacists with other healthcare professionals working in multidisciplinary teams in order to deliver optimum safety solutions for patients. It is also important to note that the usefulness of this tool may be broadened to primary care, although further studies will be needed to evaluate the tool’s performance in this setting.
The high percentage of detected ADEs caused by high-alert medications with respect to the total number of ADEs could also be explained by the characteristics of the TRIGGER-CHRON in and of itself, which includes specific signals that are associated with ADEs that occur more frequently in elderly patients with multimorbidity, and which are often caused by high-alert medications. In fact, this tool contains at least one specific trigger for each high-alert medication or drug class.
It should be noted that six of the 32 triggers included in the TRIGGER-CHRON were associated with almost 70% (148/212) of the ADEs identified on the medical records. This result is similar to findings from other trigger studies which show that just a few triggers detect the majority of potential ADEs and suggests the possibility of using a reduced number of triggers.19 20 In addition, only 25 of the 32 triggers allowed for detecting ADEs caused by high-alert medications. We believe we could use these 25 triggers, but not further limit the number of triggers, if we wish to obtain a detailed insight into the different incidents for high-alert medications occurring in elderly patients with multimorbidity. Besides, their PPVs were always >6%. In any case, this tool can be tailored to the needs of each hospital, so users can select only specific triggers focused on identifying ADEs caused by a particular drug class or individual drug when researching specific interventions focused only on those medications.
To the best of our knowledge, there are no studies available providing information on the rate of ADEs from high-alert medications in this population with which to make comparisons. However, it should be noted that the rate of ADEs found in the present study only for high-alert medications are higher than those reported in other studies that include all types of medications in older hospitalised patients,21–23 although they are close to the rates reported in studies that used triggers.24 25 These findings support the usefulness of trigger methodology for detecting ADEs and clearly show that their performance is high when a set of triggers for specific patient populations or settings are used. Finally, it should be noted that this methodology may also be used concurrently, integrating the triggers into health information technology, in order to provide real-time identification of ADEs and enable early interventions that may mitigate the ADEs detected.26
The most frequently implicated medications were corticosteroids, loop diuretics, opioid analgesics and oral anticoagulants. In this study it is only possible to make comparisons with studies that have reported medication-related harm caused by any medication in hospitalised patients in all clinical units. In these studies18 27–29 the same drug classes have been consistently implicated, although it should be noted that antibiotics and antithrombotics, which were not considered in the present study because they are not on the HAMC list, are often also responsible for a higher proportion of ADEs.
There are several limitations to our study that are inherent to the trigger methodology applied. First, retrospective review depends on the quality of the documentation on the medical records. Second, there is a certain degree of subjectivity in the review process of the medical records and in the interpretation of the triggers and ADEs, although researcher training was applied in an attempt to combat subjectivity. Third, the number of triggers is limited so they may not capture all the ADEs; however, the tool contains at least one trigger for each high-alert drug class or drug. A further limitation of the study derives from the particular high-alert medication list that was used: the Spanish HAMC list was developed for chronic patients, with the exception of corticoids that were included not only after long-term use. Our global results would have been different if we had used other lists, such as the ISMP list for hospitalised patients.29 However, we believe it is preferable to use a patient-specific list rather than a setting-specific list and that using the HAMC List for chronic patients provided us with broader, more useful information about morbidity caused by medications in this population group.
Conclusion
The application of TRIGGER-CHRON to elderly patients with multimorbidity and hospitalised in 12 Spanish hospitals allowed for the detection of one ADE caused by a high-alert drug for every four patients, and these were largely preventable. Our results confirm the need to establish interventions that reduce preventable harm with these medications. We believe that TRIGGER-CHRON can be a useful tool to measure the scope of these incidents and to determine the effects of medication safety improvement programmes as they are implemented.
What this paper adds.
What is already known on this subject
A large proportion of adverse drug events are detected in elderly chronic patients with multimorbidity.
TRIGGER-CHRON has proved to be an appropriate tool to detect adverse drug events caused in this population.
The list of high-alert medications for patients with chronic diseases (HAMC List) refers to those medications that bear a heightened risk of causing significant patient harm when used in error. For this reason, it is important to develop strategies for its prevention.
What this study adds
TRIGGER-CHRON has proved to be an appropriate tool in elderly patients with multimorbidity to detect adverse events caused by medications included in the HAMC List.
TRIGGER-CHRON could be used for measuring the effect of interventions implemented to decrease the number of incidents caused by these medications.
Acknowledgments
The authors thank the following researchers who participated in the study:
Ignacio Borbolla Día, Servicio de Medicina Interna, Hospital Ramón y Cajal, Madrid. Marta Maria Calvin Lamas, Servicio de Farmacia, Complejo Hospitalario Universitario A Coruña. Pilar Casajaús Lagranja, Servicio de Farmacia, Hospital Miguel Servet, Zaragoza. Eva Delgado Silveira, Servicio de Farmacia, Hospital Ramón y Cajal, Madrid. María José Fobelo Lozano, UGC de Farmacia, Área de Gestión Sanitaria Sur de Sevilla, Sevilla. Isabel Font Noguera, Servicio de Farmacia, Hospital Universitari i Politènic La Fe, Valencia. María García Muñoz, Servicio de Farmacia, Hospital Ramón y Cajal, Madrid. Noe Garín Escrivá, Servicio de Farmacia, Hospital de la Santa Creu i Sant Pau, Barcelona. Ana Belén Guisado Gil, UGC de Farmacia, Hospital Universitario Virgen del Rocío. Laia López Vinardell, Servicio de Farmacia, Hospital de la Santa Creu i Sant Pau, Barcelona. Laia Matas Perica, Servicio de Medicina Interna, Hospital de la Santa Creu i Sant Pau, Barcelona. María José Mauriz Montero, Servicio de Farmacia Complejo Hospitalario Universitario A Coruña. Francisca Mojer Sastre, Servicio de Medicina Interna, Hospital Comarcal de Inca, Mallorca. Esther Montero Hernández, Servicio de Medicina Interna, Hospital Universitario Puerta del Hierro-Majadahonda, Madrid. Montserrat Pérez Encinas, Servicio de Farmacia, Hospital Universitario Fundación Alcorcón, Madrid. Judit Roura Turet, Servicio de Farmacia, Hospital Clínic, Barcelona. Virginia Saavedra Quirós, Servicio de Farmacia, Hospital Universitario Puerta del Hierro-Majadahonda, Madrid. Susana Sánchez Fidalgo, UGC de Farmacia, Área de Gestión Sanitaria Sur de Sevilla, Sevilla. Sira Sanz Márquez, Servicio de Farmacia, Hospital Universitario Fundación Alcorcón, Madrid. Bernardo Santos Ramos, UGC de Farmacia, Hospital Universitario Virgen del Rocío. María Paz Revillo Pinilla, Servicio de Medicina Interna, Hospital Miguel Servet, Zaragoza. Lucia Rodríguez Cajaraville, Servicio de Farmacia, Hospital Universitario de Salamanca, Salamanca. Olga Torres Bonafont, Servicio de Geriatría, Hospital de la Santa Creu i Sant Pau, Barcelona. Montserrat Tuset Creus, Servicio de Farmacia, Hospital Clínic, Barcelona.
Footnotes
Twitter: @maritogu, @susogallego79
Contributors: All authors have contributed to the writing of the study. All authors and collaborators have contributed to the development of the project performed in the 12 Spanish hospitals.
Funding: Project “PI15/01616”, funded by Instituto de Salud Carlos III, integrated in the national I+D+i 2013-2016 and co-funded by European Union (ERDF/ESF, “Investing in your Future”).
Competing interests: None declared.
Provenance and peer review: Not commissioned; externally peer reviewed.
Data availability statement
All data relevant to the study are included in the article.
Ethics statements
Patient consent for publication
Not required.
References
- 1. U.S. Department of Health and Human Services, Office of Disease Prevention and Health Promotion . National action plan for adverse drug event prevention. Washington, DC, 2014. https://health.gov/hcq/ade-action-plan.asp [Google Scholar]
- 2. World Health Organization . WHO global patient safety challenge: medication without harm. Geneva: World Health Organization, 2017. http://apps.who.int/iris/bitstream/10665/255263/1/WHO-HIS-SDS-2017.6-eng.pdf?ua=1&ua=1/ [Google Scholar]
- 3. Sheikh A, Dhingra-Kumar N, Kelley E, et al. The third global patient safety challenge: tackling medication-related harm. Bull World Health Organ 2017;95:546. 10.2471/BLT.17.198002 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. Cohen MR, Proulx SM, Crawford SY. Survey of hospital systems and common serious medication errors. J Healthc Risk Manag 1998;18:16–27. 10.1002/jhrm.5600180104 [DOI] [PubMed] [Google Scholar]
- 5. Cohen MR, Smetzer JL, Tuohy NR, et al. High-alert medications: safeguarding against errors. In: Cohen MR, ed. Medication errors. 2nd edn. Washington, DC: American Pharmaceutical Association, 2007: 317–411. [Google Scholar]
- 6. World Health Organization . Medication safety in high-risk situations. Geneva: World Health Organization, 2019. https://apps.who.int/iris/handle/10665/325131 [Google Scholar]
- 7. World Health Organization . Medication safety in polypharmacy. Geneva: World Health Organization, 2019. https://apps.who.int/iris/bitstream/handle/10665/325454/WHO-UHC-SDS-2019.11-eng.pdf?ua=1 [Google Scholar]
- 8. Süle A, Miljković N, Polidori P, et al. Position paper on an ageing society. Eur J Hosp Pharm 2019;26:354–6. 10.1136/ejhpharm-2019-001910 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9. Avery T, Barber N, Ghaleb M, et al. Investigating the prevalence and causes of prescribing errors in general practice: the practice study. A report for the GMC, 2012. Available: www.gmc-uk.org/about/research/12996.asp [Accessed 2 Sep 2019].
- 10. Saedder EA, Lisby M, Nielsen LP, et al. Number of drugs most frequently found to be independent risk factors for serious adverse reactions: a systematic literature review. Br J Clin Pharmacol 2015;80:808–17. 10.1111/bcp.12600 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11. Toscano Guzmán MD, Galván Banqueri M, Otero MJ, et al. Validating a trigger tool for detecting adverse drug events in elderly patients with multimorbidity (TRIGGER-CHRON). J Patient Saf 2018. 10.1097/PTS.0000000000000552. [Epub ahead of print: 09 Nov 2018]. [DOI] [PubMed] [Google Scholar]
- 12. Resar RK, Rozich JD, Classen D. Methodology and rationale for the measurement of harm with trigger tools. Qual Saf Health Care 2003;12:39ii–45. 10.1136/qhc.12.suppl_2.ii39 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13. Otero MJ, Moreno-Gómez AM, Santos-Ramos B, et al. Developing a list of high-alert medications for patients with chronic diseases. Eur J Intern Med 2014;25:900–8. 10.1016/j.ejim.2014.10.021 [DOI] [PubMed] [Google Scholar]
- 14. World Health Organization . Multimorbidity: technical series on safer primary care. Geneva: World Health Organization, 2016. [Google Scholar]
- 15. Bates DW, Boyle DL, Vander Vliet MB, et al. Relationship between medication errors and adverse drug events. J Gen Intern Med 1995;10:199–205. 10.1007/BF02600255 [DOI] [PubMed] [Google Scholar]
- 16. National Coordinating Council for Medication Error Reporting and Prevention . NCC MerP index for categorizing medication errors. Available: https://www.nccmerp.org/categorizing-medication-errors-index-color [Accessed 2 Sep 2019].
- 17. Schumock GT, Thornton JP. Focusing on the preventability of adverse drug reactions. Hosp Pharm 1992;27:538. [PubMed] [Google Scholar]
- 18. Otero-López MJ, Alonso-Hernández P, Ángel Maderuelo-Fernández J, et al. Acontecimientos adversos prevenibles causados por medicamentos en pacientes hospitalizados. Medicina Clínica 2006;126:81–7. 10.1157/13083875 [DOI] [PubMed] [Google Scholar]
- 19. Singh R, McLean-Plunckett EA, Kee R, et al. Experience with a trigger tool for identifying adverse drug events among older adults in ambulatory primary care. Qual Saf Health Care 2009;18:199–204. 10.1136/qshc.2007.024406 [DOI] [PubMed] [Google Scholar]
- 20. Marcum ZA, Arbogast KL, Behrens MC, et al. Utility of an adverse drug event trigger tool in Veterans Affairs nursing facilities. Consult Pharm 2013;28:99–109. 10.4140/TCP.n.2013.99 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Vilà A, José AS, Roure C, et al. Estudio multicéntrico prospectivo de reacciones adversas a medicamentos en pacientes ancianos hospitalizados. Medicina Clínica 2003;120:613–8. 10.1016/S0025-7753(03)73788-9 [DOI] [PubMed] [Google Scholar]
- 22. Sikdar KC, Dowden J, Alaghehbandan R, et al. Adverse drug reactions in elderly hospitalized patients: a 12-year population-based retrospective cohort study. Ann Pharmacother 2012;46:960–71. 10.1345/aph.1Q529 [DOI] [PubMed] [Google Scholar]
- 23. Tangiisuran B, Davies JG, Wright JE, et al. Adverse drug reactions in a population of hospitalized very elderly patients. Drugs Aging 2012;29:669–79. 10.1007/bf03262282 [DOI] [PubMed] [Google Scholar]
- 24. Klopotowska JE, Wierenga PC, Stuijt CCM, et al. Adverse drug events in older hospitalized patients: results and reliability of a comprehensive and structured identification strategy. PLoS One 2013;8:e71045. 10.1371/journal.pone.0071045 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Lavan A, Eustace J, Dahly D, et al. Incident adverse drug reactions in geriatric inpatients: a multicentred observational study. Ther Adv Drug Saf 2018;9:13–23. 10.1177/2042098617736191 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26. Classen DC, Griffin FA, Berwick DM. Measuring patient safety in real time: an essential method for effectively improving the safety of care. Ann Intern Med 2017;167:882–3. 10.7326/M17-2202 [DOI] [PubMed] [Google Scholar]
- 27. Davies EC, Green CF, Taylor S, et al. Adverse drug reactions in hospital in-patients: a prospective analysis of 3695 patient-episodes. PLoS One 2009;4:e4439. 10.1371/journal.pone.0004439 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Pérez-Ricart A, Gea-Rodríguez E, Roca-Montañana A, et al. Integrating pharmacovigilance into the routine of pharmacy department: experience of nine years. Farm Hosp 2019;43:128–33. 10.7399/fh.11169 [DOI] [PubMed] [Google Scholar]
- 29. Institute for Safe Medication Practices . ISMP list of high-alert medications in acute care settings. Huntingdon Valley, PA: ISMP, 2018. https://www.ismp.org/sites/default/files/attachments/2018-08/highAlert2018-Acute-Final.pdf [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
All data relevant to the study are included in the article.