To our knowledge, this is the first study assessing DRPs in elderly ED patients with non-specific complaints. Our results show a prevalence of 12.2% for DRPs in this patient group. Thus, DRPs rank among the top five causes for non-specific ED presentations, with the vast majority (83%) causing acute morbidity, classified as a “serious condition”. This illustrates the importance of early detection of DRPs in the ED. In our study, 60% of DRPs in patients visiting the ED with NSC were initially not diagnosed as medication-related, and were also not considered in the differential diagnosis.
The risk for a DRP increased with the number of prescribed medications and with treatment with certain drug classes, in particular antidepressants, benzodiazepines and anticonvulsants. Intake of thiazides, benzodiazepines, antidepressants and anticonvulsants was associated with a significantly increased risk for DRPs. The corresponding final diagnosis was most commonly hyponatremia or medication-overdose.
Non-specific complaints are a common mode of presentation in the ED and have previously been described using terms, such as “general deterioration”, “loss of energy”, “weakness” or “home care impossible” [19, 20, 34, 45]. Elderly patients with comorbidities, who are not institutionalized, belong to the high-risk population developing DRPs [4, 12, 46]. This risk group is highly represented in our study population. The vast majority of our patients with NSCs were not institutionalized (89.6%) and had a relatively high burden of comorbidity compared to similar populations described [47, 48].
It has been previously shown that for an elderly ED population the most common drugs causing DRPs are diuretics, oral anticoagulants, NSAIDs, antiarrhythmics, antiplatelet agents and psychotropic drugs . In another study on older patients, Warfarin, Insulin, oral antiplatelet agents and oral hypoglycemic agents were implicated alone or in combination in 67% of emergency hospitalizations for adverse drug events . However, drug classes such as anticoagulants, NSAIDs, antiarrhythmics, and antiplatelet drugs tend to cause specific symptoms (e.g. bleeding, syncope) or signs (e.g. bradycardia, hypotension) in contrast to diuretics. This may explain the under-representation of DRPs induced by these substances in our study population whose chief complaints were merely non-specific.
In accordance with the Beers criteria, we observed several cases of DRPs due to SSRIs, benzodiazepines and anticonvulsants . However we also detected DRPs associated with drugs which are neither listed in the Beers criteria of 2003 nor 2012, e.g. thiazide-diuretics . This discrepancy which was also observed in other studies [51–53], could be due to a different prescribing pattern or differing opinions about inappropriateness between the US and Europe [54, 55]. Another publication compared seven explicit criteria of drug inappropriateness in elderly patients from different countries, including the Beers criteria . The only drugs considered inappropriate by all seven criteria were long-acting benzodiazepines and tricyclic antidepressants . Diuretics were listed by four of the seven criteria, mainly as potentially inappropriate in combination with other drugs (e.g. NSAIDs, digoxin) or in patients with a history of gout [57–60].
Importantly, one study revealed that up to 40% of drug-related ED visits are not correctly diagnosed by Emergency Physicians (EP) in a general ED population . Obviously, DRPs can be difficult to diagnose for EPs, especially when elderly patients present with non-specific complaints. In this subgroup of patients, the proportion of initially missed DRPs was even higher in our study. A potential reason for this might be that EPs are better at identifying DRPs which relate to the patients’ chief complaints as compared to DRPs with symptoms unrelated to their chief complaints . However, NSCs were not specifically addressed in that study. The high proportion of DRPs that were not identified in our study strongly supports the hypothesis that non-specific manifestations of DRPs are more likely to be missed on admission than DRPs associated with specific symptoms.