Elderly patients with inappropriate medication correlations with adverse drug events or unexpected illnesses in long-term care institutions

Abstract

Introduction: The elderly are subject to natural aging and the health problems caused by the recession of physical and mental functions. Elderly patients are also more susceptible to adverse reactions of medication, drug interactions and other drug problems than the young. We then investigated patients with adverse drug events (ADEs) or unexpected illnesses transferred to our hospital during the whole year of 2010.

Methods: We analyzed the medicine of elderly patients in long-term care institutions. Four long-term care institutions with different types and sizes located near Fong-Yuan Hospital in downtown Fong-Yuan were investigated. In this study, the researchers divided potentially inappropriate medications (PIMs) into two categories: (a) those with or without the drug–drug interaction (DDI) and (b) those with narrow therapeutic index drugs. Variables were reclassified as inferential statistics for analysis by using the independent t-test or Mantel–Haenszel test.

Results: The data for age, gender, presence or absence of dementia, brain damage and Parkinson’s disease were divided into two groups for those patients with or without PIMs. There were no statistically significant differences among the groups. However, the numbers of chronic diseases for the group with PIMs were higher, and the numbers of drug items with PIMs were also higher. In addition, we investigated the presence or absence of PIMs for patients transferred to our hospital with ADEs and unexpected illnesses. The results showed no statistically significant differences among the groups.

Conclusions: Our results showed that elderly patients who had consultations with doctors and the hidden problems about medication were detected by pharmacists in the privileged hospital had no direct risk with DDI or narrow therapeutic index drugs. However, other potential drug risks remain to be further analyzed and more samples should be surveyed.

• Adverse drug event
• inappropriate medication
• long-term care institution
• unexpected illness

Introduction

In Taiwan, the aging of the population is increasing at an alarming rate. The Department of Health, Executive Yuan, R.O.C, stated that the proportion of the elderly population was 6.52% in 1991 and more than 7% in September 1993. Thus, Taiwan crossed the threshold of an aging society according to the definition of the World Health Organization. The elderly population continued to increase to 9.23% by the end of 2003 and reached approximately 208 million people.

The elderly are subject to natural aging and the health problems caused by the recession of physical and mental functions. Moreover, the risk with a variety of chronic diseases and drugs, older patients are more likely to suffer from polypharmacy. Elderly patients are also more susceptible to adverse reactions of medication, drug interactions and other drug problems than the young.

Elderly patients accounted for one-fifth of outpatient visits. Percentages of patients 65–74 years of age and those over the age of 75 years reporting two or more chronic diseases were 59.7% and 68.7%, respectively, with 2.36 and 2.75 diseases per elderly patient [1]. A number of problems often occur in the medication of the elderly, including self-withdrawal, the use of non-prescription drugs, the use of a variety of drugs, taking of donated drug and forgetting to take medicine. Peck C.L. reported that only 22% of medicine was taken correctly, and 31% were being misused in a manner that posed a serious threat to the patient’s health [2].

The knowledge about medication of the elderly, which often comes from their own experiences and adverse hearsay, directly affects the drug safety of older patients. Pharmacists provide pharmaceutical health education in long-term care facilities, consulting services and medication guides to communicate directly with elderly patients or family care providers. Pharmacists also assist front-line nurses or care staff in improving the quality of pharmaceutical care. So it should be possible to develop a more effective system for improving the medication environment and quality.

Methods

We analyzed the medicine of elderly patients in long-term care institutions. Four long-term care institutions with different types and sizes located near Fong-Yuan Hospital in downtown Fong-Yuan were investigated. Our hospital, a sizable hospital in Fong-Yuan, is the primary and privileged hospital for these four institutions. The hospital pharmacists check prescriptions with or without potentially inappropriate medications (PIMs) and give recommendations as references for the prescribing physician. We analyzed the appropriateness of the medications, possible drug–drug interactions (DDIs) and side effects to explore the risk factors associated with PIMs during the whole year of 2010.

The purpose of this study was to explore elderly patients taking outpatient prescription drugs. Chinese medicine, dental and emergency prescriptions were not included in this analysis. Exclusion criteria were terminal illness and a life expectancy of less than 6 months; expected length of stay care institution less than 3 months; subjects without any prescription coverage were excluded. PIMs criteria were based on the DDI system [3] and narrow therapeutic index drugs [4]. DDIs occur when the effect of one drug is changed by the presence of another drug. The outcome can be harmful if the DDI causes an increased toxicity of the drug. However, a reduction in therapeutic efficacy due to a DDI may be just as harmful as an increase. Drugs have a narrow therapeutic index and require regular monitoring to prevent overdose or toxicity, e.g. warfarin and digoxin. Underuse of appropriate medications is common among elderly persons.

In this study, the researchers divided PIMs into two categories: (a) those with or without the DDI and (b) those with narrow therapeutic index drugs. We then investigated elderly patients had taken outpatient prescription drugs over a long-term due to chronic diseases and they had adverse drug events (ADEs) or unexpected illnesses transferred to our hospital in 2010. Unexpected illnesses were identified to include admission to emergency department (ED) or hospitalization in our hospital. ADEs were identified to have occurred when documentation was provided by the physician in the medical records suggesting signs or symptoms that may have been drug related. An ADE refers to any injury caused by the drug (at normal dosage and/or due to overdose) and any harm associated with the use of the drug.

This study used the Microsoft SQL 2003 required processing database for file concatenation, variable capture and information to establish the necessary data files. Statistical analysis was carried out by SPSS version 13 (SPSS Inc., Chicago, IL). Descriptive statistics included frequency analysis of the variables, percentage, mean and standard deviation. Variables were reclassified as inferential statistics for analysis by using the independent t test or Mantel–Haenszel test. In this study, all variables with p values less than 0.05 were defined as significant difference.

Results

Elderly patient characteristics

There were 184 elderly patients in this analysis. Sixty patients were aged 65–69 years (32.6%), 100 patients suffered from one to three chronic diseases (54.3%), 22 patients had dementia (12.0%), 7 patients had Parkinson’s disease (3.8%) and 33 patients had other central nervous system (CNS) disorder (17.9%) (see Table 1).

Table 1. Baseline characteristics of the study (n = 184).

Age (years)	n	Percent
65–69	60	32.6
70–74	48	26.1
75–79	38	20.7
≧80	38	20.7
Gender
Male	97	44.5
Female	87	55.5
Number of chronic disease  per patient
1–3	100	54.3
4–6	84	45.7
Dementia diagnosis
Yes	22	12.0
No	162	88.0
Parkinson’s disease diagonsis
Yes	7	3.8
No	177	96.2
Other CNS disorder (e.g. stroke  residua, TIA, mental disorders,  degenerative diseases)
Yes	33	17.9
No	151	82.1
Number of drug per prescription
1–4	38	20.8
5–8	115	62.5
≧9	31	16.8
Drug–drug interaction (grade)
0	133	72.3
1–3	51	27.7
Narrow therapeutic index drug
Digoxin	9	4.9
Carbamazepine	3	1.6
Cyclosporine	0	0
Lithium	0	0
Phenytoin	14	7.6
Theophylline	23	12.5
Warfarin	4	2.2
Patients with ADE
Yes	3	1.6
No	181	98.4
≧1 admissions in the ED
Yes	19	10.3
No	165	89.7
≧1 hospital admissions
Yes	63	34.2
No	121	65.8

ADE: adverse drug event; CNS: central nervous system; ED: emergency department; TIA: transient ischemic attack.

The greatest number of patients, 115, had 5–8 drugs per prescription (62.5%) and 51 patients had grade 1–3 DDI (27.7%). Three patients with ADEs entered our hospital for treatment during the whole year. Nineteen patients with unexpected diseases entered our ED (see Table 2).

Table 2. Comparison of patients exposed and unexposed to drug–drug interaction.

	Drug–drug interaction
	Yes	No	p Value
Age	73.16 ± 14.29	73.14 ± 14.49	0.995
Gender (male/female)	22/25	74/63	0.198
Number of chronic  disease per patient	3.64 ± 1.0	3.21 ± 1.36	0.023
Dementia diagnosis  (yes/no)	3/42	19/120	0.096
Parkinson's disease  diagnosis (yes/no)	2/43	5/134	0.592
Other CNS disorder (yes/no)	6/39	27/112	0.284
Number of drug per prescription	7.49 ± 1.6	6.0 ± 2.44	0.0002
Patients with ADE
Odds ratio 95% CI	6.233 (0.551–70.436)		0.315
≧1 admissions in the ED
Odds ratio 95% CI	0.951 (0.39–2.291)		0.912
≧1 hospital admissions
Odds ratio 95% CI	1.103 (0.505–2.41)		0.963

ADE: adverse drug event; CNS: central nervous system; ED: emergency department.

PIMs were divided into two major categories based on analysis results.

• Those with or without DDI: In this category, data were divided into two groups, those with or without the DDI, based on age, gender, presence or absence of dementia, brain damage, Parkinson’s disease and other CNS disorder. There were no statistically significant differences between the two groups. The numbers of chronic diseases per patient (3.64 ± 1.0 versus 3.21 ± 1.36, p = 0.023) and drugs per prescription (7.49 ± 1.6 versus 6.0 ± 2.44, p = 0.0002) with DDI were higher than those without.

We investigated the patients transferred to our hospital because of ADEs (odds ratio (OR): 6.233, 95% CI: 0.551–70.436, p = 0.315) and those who had unexpected illness admissions to ED (OR: 0.951, 95% CI: 0.39–2.291, p = 0.912) or hospitalization (OR: 1.103, 95% CI: 0.505–2.41, p = 0.963) in our hospital. There were no statistically significant differences between the two groups.

• Those with or without narrow therapeutic index drugs: In this category, data were divided into two groups, those with or without narrow therapeutic index drugs, based on age, gender, presence or absence of dementia, brain damage, Parkinson’s disease and other CNS disorder. There were no statistically significant differences between the two groups. The numbers of chronic diseases per patient (3.74 ± 0.9 versus 3.19 ± 1.36, p = 0.003) and drugs per prescription (7.36 ± 1.48 versus 6.07 ± 2.46, p = 0.002) with narrow therapeutic index drugs were higher than those without (Table 3).

Table 3. Comparison of patients exposed and unexposed to narrow therapeutic index drug.

	Narrow therapeutic index drug
	Yes	No	p Value
Age	72.79 ± 13.09	73.25 ± 14.73	0.846
Gender (male/female)	25/18	69/69	0.281
Number of chronic  disease per patient	3.74 ± 0.9	3.19 ± 1.36	0.003
Dementia diagnosis  (yes/no)	3/39	19/123	0.147
Parkinson's disease  diagnosis (yes/no)	1/41	6/136	0.507
Other CNS disorder  (yes/no)	9/33	24/118	0.576
Number of drug per prescription	7.36 ± 1.48	6.07 ± 2.46	0.002
Patients with ADE
Odds ratio 95% CI	6.85 (0.605–77.512)		0.272
≧1 admissions  in the ED
Odds ratio 95% CI	1.064 (0.44–2.573)		0.928
≧1 hospital admissions
Odds ratio 95% CI	0.94 (0.438–2.018)		0.972

ADE: adverse drug event; CNS: central nervous system; ED: emergency department.

We investigated the patients transferred to our hospital because of ADEs (OR: 6.85, 95% CI: 0.605–77.512, p = 0.272) and those who had unexpected illness admissions to ED (OR: 1.064, 95% CI: 0.44–2.573, p = 0.928) or hospitalization (OR: 0.94, 95% CI: 0.438–2.018, p = 0.972) in our hospital. There were no statistically significant differences between the two groups.

Discussion

Improper medication-related risk factors for the elderly include many types of medication (medication ≥ 5 kinds) [5], doctor shopping and polypharmacy. Many studies have pointed out that early or immediate discovery of drug problems. It is necessary for patients to avoid unnecessary harm. Moreover, safety assessment of inappropriate drugs should be conducted to ensure the quality of care and patient medication.

Studies have stated that medication errors (the wrong drugs, the wrong time, the wrong patient, etc.) and execution errors (such as wrong infusion rate) are common in long-term care [6]. Annual statements from Taiwan’s patient safety notification system reported (from January 2005 to 2009) on 31 December 2010. The top five on adverse healthcare events included fall events, drug events, the piping incident events, medical care events and hurtful behavior events. In 2009, the top three events were drug events, fall events and piping events, respectively [7]. Atiqi R. mentioned that ADEs accounted for 3.4–33.9% of all the reasons leading to hospitalization [8].

Of 528 patients, 103 (19.5%) were admitted for iatrogenic events; drug adverse events accounted for 49.0% of iatrogenic events [9]. If we are unable to improve the correctness of administration, we will not be able to reduce the incidence of ADEs and the resulting invalid treatment [10]. For the above reasons, we should help care institutions to improve their administrative correctness to medication safe for patients.

The four institutions are located near Fong-Yuan Hospital in downtown Fong-Yuan.

Those patients had consultations with doctors and the hidden problems about medication were detected by pharmacists in the privileged hospital.

In 1984, Klein L.E. et al. [11] found that 25.2% older patients taking medication reported adverse symptoms with a least one drug. Patients could take actions in response to medication-related adverse symptoms. They took more or less different medication (7.4%) and discontinued (4.8%). This finding shows that older patients reported fewer side effects than younger. First, it is possible that the elderly are not more susceptible to adverse drug reactions. Second, the elderly do experience more medication side effects, but are not aware of them. This leads to a low prevalence of PIMs.

An observational study was to examine the prevalence of PIMs and ADEs in older adults presenting to the ED. PIMs were defined according to the 2003 Beers criteria. Forty-four potential adverse drug interactions were identified in 33 patients (26.6%; 95% CI: 19–34%); 9 patients (27% of those with ADEs) had multiple potential ADEs [12].

Study of Hutchinson et al. [13] was published in 1986, they used Naranjo’s algorithm to assess the suspected reactions for drug causation with the following results: 1% definite; 19% probable; 66% possible; and 14% unlikely. The rate of probable or definite reactions was 49/1026 (5%) per patient and 58/3330 (2%) per drug course. Our study showed three patients (1.6%) transferred to our hospital because of ADEs for treatment, and ADEs criteria were relatively strict, so the result was much lower prevalence than Klein L.E. et al.’s study in ADEs.

Study of Zermansky et al. [14] reported that a clinical pharmacist can review prescriptions and make recommendations that were usually accepted. This leads to substantial change in patients’ medication regimens without change in drug costs. There was a reduction in the number of falls. However, there was no significant change in consultations, hospitalization and mortality. Our results showed no statistically significant differences in hospitalization and ED admission.

Conclusion

It is very important to consider elderly patients in pharmaceutical care, medication safety and drug use. Our results showed that elderly patients who had consultations with doctors and the hidden problems about medication were detected by pharmacists in the privileged hospital had no direct risk for DDI or narrow therapeutic index drugs. However, other potential drug risks remain to be further analyzed, and more samples should be surveyed. Increased relative risk of the older population in association with inappropriate medication or contraindications of drug use deserves careful evaluation by clinical caregivers. It is important to confront the issues of medical care in order to improve the level and quality of medication used in the care of the elderly.

Declaration of interest

The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.