Prevention of medication errors at hospital admission: a single-centre experience in elderly admitted to internal medicine

Abstract

Background Transition of care on admission to the hospital and between clinical areas are risk points for medication errors. All type of medication errors can be reduced by improving communication at each transition point of care. Objectives This study examines the impact of pharmacist obtained best possible medication histories on medication errors at admission due to unintentional medication discrepancies in older patients. Setting This was a prospective, single-center study conducted in an Internal Medicine Department of a tertiary care teaching hospital in Saudi Arabia. Methods Patients ≥ 65 years with an existing drug therapy on admission were eligible. The best possible medication history taken by the pharmacist from different sources of medication information was compared to the admission medication order to identify and correct unintentional discrepancies. The discrepancies were classified according to the type of errors. An independent multidisciplinary team adjudicated the potential for harm of each type of medication error. Main outcome measure Number and proportion of unintentional medication discrepancies upon admission and associated medication errors. Secondary outcomes included clinical significance and drug classes involved in the discrepancies and risk factors for the occurrence of these discrepancies. Results A total of 375 evaluable patients were identified. Among 375 medication histories, 609 discrepancies were detected of which 226 were recorded as unintentional. 151 patients (42.4%) had ≥ 1 unintended discrepancy. Drug omission (37%) was the most frequent type of error. Nervous system (24.5%), and cardiovascular system (21.2%) were the most common drug classes involved in medication errors. Three-fifths of the UMD had the potential to cause temporary harm with initial or prolonged hospitalization. The number of medications prescribed upon admission (OR 1.32, 95% CI 1.09–1.54, p < 0.034), number of sources consulted for the best possible medication history (OR 1.53, 95% CI 1.38–1.76, p < 0.01) and the completion of medication review process within 24 h (OR 0.89, 95% CI 0.86–0.94, p < 0.03) of the admission were the 3 most significant predictors of the discrepancies. Conclusions In elderly patients, medication histories are often recorded inaccurately by physicians at the time of hospital admission, this creates the potential for medication errors starting at admission. In older adults, best possible medication histories are also useful in detecting drug related pathology or drug–drug interactions.

Appendix

Examples of medication errors detected in study according to type of unintentional medication discrepancies

Type of error	Examples
Drug omission	(1) The omission of antifungal prophylaxis in hematopoietic stem cell transplant patient (2) The omission of phenytoin in a patient with a recent new onset of seizure following neurosurgery (3) The omission of vancomycin coverage in a patient with confirmed MRSA infection (4) The omission of prophylaxis against Pneumocystis carinii Pneumonia in the immuno-compromised patient (5) The omission of ibuprofen in a patient recently started on allopurinol for gout (6) The omission of warfarin at admission in a patient with mitral valve replacement surgery
Discrepant or incorrect dose	(1) The physician prescribes “warfarin 10 mg” on admission while “simvastatin 10 mg” was recorded in patient home medication list that was continued upon admission without dose adjustment. Simvastatin may enhance the anticoagulant effect of warfarin (a clinically significant interaction) (2) “Omeprazole 20 mg PO” was prescribed upon admission, while dabigatran was recorded as a home medication. Omeprazole may decrease serum concentrations of the active metabolite of Dabigatran. Monitor drug therapy is recommended (3) “Wellbutrin XL (bupropion) 300 mg OD” was prescribed upon admission, while “Wellbutrin SR (bupropion) 150 mg” was recorded as a home medication. An abrupt increase in dose may cause Insomnia and agitation (4) Controlled-release carbidopa/levodopa 50/200 (Sinemet CR) twice daily was prescribed upon admission, while standard carbidopa/levodopa (Sinemet 25/100) once daily was recorded as a home medication for restless legs syndrome. The discrepant dose was prescribed for Parkinson disease instead of restless legs syndrome (5) “Phenytoin 100 mg TID” for post-hemorrhagic stroke seizures as home medication was continued on admission, while “Invokana (canagliflozin) 100 mg” was prescribed on admission as an add-on treatment for type-2-diabetes mellitus. Phenytoin may decrease the serum concentration of canagliflozin and it is recommended to start treatment at 300 mg/day in patients receiving concomitant phenytoin
Discrepant or wrong frequency	(1) “Terlipressin acetate 2 mg/4 h IV infusion” for bleeding oesophagal varices associated with liver cirrhosis was recorded in admission medication order, while “terlipressin acetate 2 mg/4 h IV bolus” was noted inter-hospital transfer records. Evidence for the use of terlipressin by continuous intravenous infusion is not convincing and it is currently not approved (2) Patient with a history of NYHA Class I CHF prescribed “pioglitazone 15 mg BID”, while pioglitazone 15 mg OD was recorded in patient home medication list. There was a two-fold increase in the dose of pioglitazone than recommended dose in patients with NYHA Class I or Class II CHF (3) “Isosorbide dinitrate ER 40 mg TID” was prescribed on admission, while “isosorbide dinitrate 40 mg BID” was recorded in pre-admission medications. An abrupt increase in dose due to change in tablet formulation (4) “Pregabalin 300 mg/day TID” for diabetes-associated neuropathic pain was recorded in admission, while “pregabalin 75 mg BID” was recorded as a home medication. Pregabalin dosage greater than 500 mg/day increase risk of adverse effect with no additional benefit (5) “Carvedilol ER 80 mg OD” for CHF was recorded in pre-admission medications, while “carvedilol IR 3.125 mg BID” was recorded in admission medications. The physician ordered starting dose of carvedilol for the treatment of CHF on admission, however, the patient was already on escalated-dose of carvedilol for post-MI heart rate control
Drug commission (previously stopped drug)	(1) A patient recently started on “sertraline 50 mg” for social anxiety disorder was admitted for acute diarrhoea. The physician failed to discontinue sertraline. Sertraline tends to cause more loose stools than constipation and should likely be avoided over other alternatives that would be less likely to exacerbate diarrhoea (2) Sitagliptin was recorded in a patient admitted for suspected acute pancreatitis. The treating physician missed discontinuing sitagliptin as a possible suspect or interacting agent of patient condition (3) A patient admitted for an acute urinary tract infection with delirium was taking “oxybutynin” before admission. Oxybutynin was also recorded in admission medication order. The treating physician failed to identify oxybutynin as a precipitating risk factor for delirium (4) Patient with a history of CHF and HTN admitted for the management of acute renal colic, was taking “hydrochlorothiazide 50 mg daily” before admission. Hydrochlorothiazide 50 mg was also recorded in admission order. Based on patients’ renal function SCr 3.1 mg/dL, eGFR 26 mL/min/1.73 m2, hydrochlorothiazide should be held or substituted with another drug (5) “Moxifloxacin 400 mg IV OD” was recorded in patient’s admission order for the management of COPD exacerbation. However, past medical record showed that the patient reportedly discontinued levofloxacin due to tendinitis
Incorrect drug	(1) The patient was taking “tiagabine” for seizure control, however, upon admission, he was prescribed “tizanidine” (2) “Metoclopramide 10 mg IV” was prescribed for acute diabetic gastric stasis. However, the patient record shows taking “risperidone 2 mg” in the past 6 months for major depression. Concomitant use of risperidone and metoclopramide more likely to cause tardive dyskinesia (3) A patient with a medical history of CHF was admitted for acute gastroenteritis with severe nausea, vomiting and hypokalemia. “Ondansetron 10 mg IV” was prescribed upon admission for the treatment of vomiting. Ondansetron is a QTc-prolonging agent. Hypokalemia is an additive risk factor for drug-induced QT prolongation, particularly in patients with CHF (4) “Atorvastatin 10 mg” substituted for “rosuvastatin” for secondary prevention stroke, however, past medical record showed that the patient recently switched to rosuvastatin due to complain of distal Muscle weakness (5) “Celecoxib 100 mg BID” was prescribed for an acute gout episode to a patient with a recent history of unstable angina. NSAIDs, in particular, COX-II inhibitors, may pose a risk of serious cardiovascular thrombotic events. This especially important in a patient with cardiovascular disease

1. PO per oral, XL extended-release, SR sustained-release, TID three times a day, NYHA New York Health Association, CHF congestive heart failure, ER extended-release, IR immediate release, MI myocardial infarction, HTN hypertension, SCr serum creatine, eGFR estimated glomerular filtration rate, COPD chronic obstructive pulmonary disease, NSAIDs non-steroidal anti-inflammatory drugs, COX-II cyclo-oxygenase II, IV intravenous, MRSA methicillin-resistant Staphylococcus aureus