Polypharmacy in the oldest old (≥80 years of age) patients in China: a cross-sectional study

This was a cross-sectional study (convenience sample) of 258 oldest old patients (≥ 80 years of age) and hospitalized at a tertiary hospital in Beijing between December 1, 2014 and June 30, 2015. Their pre-admission status was assessed using three validated questionnaires. The study was approved by the ethics committee of the hospital. All patients provided a written informed consent.

The inclusion criteria were: 1) ≥80 years of age; 2) at least one chronic disease, selected basing on “China Report of the Development on Aging Cause Blue Book” [1] and “Report of Status and Trend in Development of Chronic Diseases in Older Adults in China”, which was published by China Research Center on Aging (hypertension, coronary heart disease, diabetes mellitus, hyperlipidemia, chronic gastritis, chronic low back pain, atherosclerosis, respiratory disease, or skin or tissue disease); and 3) taking ≥5 types of medications before admission (including prescription medications but not including non-prescription medications or over-the-counter health care products). The exclusion criteria were: 1) life-threatening or terminal disease; 2) mental disorder or cognitive impairment; 3) or unwilling to participate.

Following an explanation of the purpose of the study, consent was obtained from patients who met the eligibility criteria and agreed to participate. Demographic and medical data were obtained from the patients and their medical records. The three questionnaires took approximately 20 min to fill and the patients filled the questionnaires at the hospital. Pill boxes were provided as an incentive for participation. Before the patients completed the questionnaires, the meaning and requirement of all items were explained in details by a healthcare professional. The questionnaires were filled by the patients, without help, and in the presence of the professional. All questionnaires covered the pre-admission period.

The general demographic questionnaire included sex, age, educational level, marital status, and payment of medical expenses. The subjects also provided a self-evaluation of their health status, the types and number of diseases they had, and the number and names of the drugs they were taking. This information was validated using the patient’s medical records. For analysis convenience, medications were recorded and coded according to “Practical Drug Code for Hospital Pharmacy” which is occupation standard of medicine in China and used in hundreds of medical organizations, compiled and published by Pharmacy Department of Peking Union Medical College Hospital.. History of adverse reactions (ADR) was obtained from the patients’ medical records and the ADR Monitoring System of the hospital. ADR monitoring system was based on Hospital Information System (HIS) System of internal network in Peking Union Medical College Hospital. Electronic system simplified the procedure of reporting and recorded all the ADRs roundly. The reporter could inquire about the content of the report and consult the feedback content in time by the system.

The medication knowledge questionnaire was developed by the authors based on an extensive literature review of Chinese and English papers [20‐27], relevant work experience, and suggestions from five geriatrists and pharmacists. The resulting questionnaire was revised by a group of faculty members from a hospital in Beijing. The content validity was evaluated by five geriatrists and pharmacists. A strong positive correlation was defined as 4 points, positive correlation as 3 points, weak correlation as 2 points, and no correlation as 1 points. The content validity index (CVI) of the questionnaire was 0.96, and the CVI of the items ranged from 0.71–1.00. The Cronbach’s α was calculated as a measure of the internal validity of the questionnaire and was 0.82. The questionnaire (English translation, non-validated) is presented as Additional file 1. The questionnaire was also assessed for ease of comprehension and readability by 15 oldest old patients with polypharmacy (not participating in the present study).

This questionnaire included eight dimensions and 25 questions. The content of the items addressed the subjects’ knowledge about: 1) medication delivery modes; 2) medication dosage; 3) the importance of taking medications regularly; 4) drug side effects; 5) the dangers of drug abuse (self-medication without a doctor’s guidance and/or beyond the medical scope of treatment and dose standards); 6) how to identify expired drugs; 7) drug incompatibilities; and 8) drug storage. The average awareness level was calculated as the number of questions answered correctly divided by the total number of questions.

The 4-item Morisky scale was used to assess medication compliance [28]. Xu et al. [29] evaluated the internal reliability of the Chinese version of the Morisky scale in Chinese patients and reported a Cronbach’s α of 0.715. The scale consists of four questions regarding the 6 months prior to admission: 1) Have you ever forgotten to take your medicine? 2) Are you careless about taking your medicine? 3) Do you sometimes stop taking your medicine when you feel better? 4) Sometimes, if you feel worse when you take your medicine, do you stop taking it? The patients were considered adherent if their answers to all four questions were “No”, and they were considered poorly adherent if one or more of their responses was “Yes”. The rate of medication adherence was defined as the number of compliant subjects divided by the total number of subjects.

The 2012 American Geriatrics Society’s updated Beers criteria [30] were used to assess PIMs. The new 2012 Beers criteria were translated into Chinese and adapted in accordance with the National Essential Drugs List of China (2009 edition), omitting medicines unapproved in China [31]. PIMs were defined as drugs with a potential for harm that outweighed their benefits. The medication review chart was used to assess PIMs comprehensively, and the medication review chart (English translation, non-validated) is presented as Additional file 2. In this study, PIM was evaluated by five attending physicians (all had medical doctor degree, and worked in geriatric wards as attending physicians at least 5 years). At least one inappropriate medication in the patient’s prescription regimen, as based on the 2012 Beers criteria, was considered to indicate the use of PIMs.

Of the 258 subjects, the age range was from 80~ 109; mean age was 88.7 ± 6.5 years. Among the subjects, males represented 58.9%. The majority (55.8%) of the patients lived with their spouses and 96.9% of the subjects received free medical care or were covered by health insurance. The majority of patients (53.9%) had junior college diplomas, and 24.7% had middle-school education or below (Table 1).

The number of chronic diseases was 3–13/subject (mean of 7.0 ± 2.3). The five most common chronic diseases were hypertension (n = 160, 62.0%), hyperlipidemia (n = 112, 43.4%), atherosclerosis (n = 111, 43.0%), chronic gastritis (n = 106, 41.4%), and coronary heart disease (n = 105, 40.7%) (Table 2).

The five most commonly used prescription drugs were: gastrointestinal drugs (n = 199), vitamins and minerals (n = 185), traditional Chinese medications (TCM, refer particularly to Chinese patent medicines, which were produced by traditional Chinese herbal medicine according to the formulary prescription and preparation process, and had definite and systemic effects.) (n = 156), osteoporosis prevention drugs (n = 143), and antithrombotic drugs (n = 142) (Table 3).

The maximum number of types of medications used by the oldest old with polypharmacy was 27. The majority of the subjects (55.4%) took < 10 types of drugs. Patients taking 11–20 pills/tablets of drugs accounted for 46.1% of the patients. Subjects with a history of adverse drug reactions accounted for 40.3%. In addition, 60.5% of the subjects took Chinese traditional medicines, and 64.7% of the subjects used over-the-counter health care products (Table 4).

The average accuracy of the subjects’ responses to the medication knowledge items was 68.7%. The items with the highest average numbers of correct responses included those on expired drugs, medication types, and drug storage, whereas the items with the lowest awareness were about drug side effects, drug abuse, drug incompatibility, and medication dosage (Table 5). Correlations were identified between age and medication knowledge; medication types and medication knowledge; and numbers of drugs and medication knowledge. Medication knowledge declined with increasing age: for each one-unit increase in age, the average medication knowledge score decreased by 0.697 (r = − 0.697, P < 0.05).

The proportion of PIMs prescribed to oldest old patients with polypharmacy was 27.1%. Based on the prescriptions of the 258 subjects, there were a total of 70 PIMs prescribed for 61 patients. The specific distribution is shown in Table 6.

In this study, only 32.6% of the oldest old patients with polypharmacy were compliant to their treatment regimen. Of the four items on the Morisky scale, compliance was the lowest (50.0%) for forgetting to take medicines and the highest (59.3%) for sometimes being careless about taking medicine.

Ten variables (gender, age, education level, marital status, medical costs, disease diagnosis, medication classes, quantity of drugs, adverse drug reactions, and medication knowledge) were analyzed using univariate and multivariate logistic regression analyses to determine the factors associated with compliance. The univariate logistic regression analysis showed that gender (P < 0.001), age (P < 0.001), marital status (P = 0.001), medical expenses (P = 0.003), disease diagnosis (P < 0.001), medication classes (P = 0.001), number of drugs (P = 0.002), and medication knowledge (P < 0.001) were associated with compliance. These variables were entered in the multivariate logistic regression analysis, revealing that age and medication classes were independently negatively associated with compliance, and that medication knowledge was independently positively associated with compliance (Table 7).

In the present study, oldest old of ≥ 80 years of age were specifically selected because they represent a relatively novel and expanding population in China’s aging society [32]. The majority of the patients in our study lived with their spouses, which is consistent with Chinese national data and the social characteristics of oldest old living arrangements [32]. This suggests that most of the subjects did not lack for care and had some form of physical and psychological support. Regarding health care costs, 96.8% of the subjects received free medical care or had health insurance, suggesting that the patients’ economic burden was light and that few of them were concerned about their medical expenses. The average number of chronic diseases per patient observed in this study was generally consistent with previous studies in China [33, 34]. The most commonly used prescription drugs were gastrointestinal drugs. Classification statistics was conducted in the cardiovascular drugs as follow: calcium antagonists, beta blockers, thiazide drugs, angiotensin, antithrombotic drugs and statins. So the cardiovascular drugs were very commonly used in older adults if aggregated in this study..

Previous studies showed that appropriate medication knowledge was present in 60–72% of people [35‐37]. In the present study, the subjects specifically lacked knowledge about drug side effects, drug abuse, and medication dosage, as supported by previous studies worldwide [38, 39]. These findings may be related to the participants’ advanced age, as the regression analysis results showed that medication knowledge declined with increasing age, which is supported by previous studies in different populations worldwide [40‐42]. Previous studies showed that medication knowledge is intimately associated with compliance [35, 37, 39, 43, 44], but since multiple factors are involved in compliance, medication knowledge is thought to play a relatively minor role [45]. In this study, the rate of medication compliance among the oldest old patients with polypharmacy was lower than in previous studies [14, 15, 18, 39, 41, 46, 47]. Patients may forget to take their drugs because of poor self-management, poor memory, or a limited emphasis on taking routine medication [48]. It is known that medication compliance among the older adults is affected by many factors, including patient-related factors, social support, medical expenses, and the doctor-patient relationship [14, 15, 18, 39, 41, 46‐48]. Although certain social demographic characteristics (e.g. age and gender) and disease factors cannot be changed, several patient-related factors (e.g., medication knowledge level) can be intervened on, improving compliance. This study showed that age and medication classes were independently associated medication compliance, and that medication knowledge was an independent protective factor, as supported by previous studies [48‐50], and could be more important to medication compliance than previously suggested [45]; nevertheless, the present study examined Chinese patients ≥ 80 years old, while Lam et al. [45] examined Australian patients ≥ 65 years of age.

The use of an excessive number of types of drugs can increase the risk of adverse drug reactions and the complexity of dosing (e.g., the method of administration and frequency of medication use). The numbers of drugs used in the present study were higher than those reported in previous studies [12, 51, 52]. Traditional Chinese medicine must be considered since it is generally not used outside China, is counted when evaluating polypharmacy, and can lead to adverse drug reactions [53]. In the present study, the proportion of elderly patients taking traditional Chinese medicine was high, highlighting the potentially high impact of traditional Chinese medicine on polypharmacy in Chinese populations. In addition, the proportion of oldest old patients using over-the-counter health care products in our study was high, which was likely related to the light economic burden of the subjects. The numbers of drugs may influence patients’ compliance to treatment. Indeed, patients may worry that excessive drug use could lead to adverse reactions, which may affect compliance, as supported by the fact that 40.3% of the subjects in our study had a history of adverse drug reactions, which was higher than that of previous studies [47, 54, 55]. Therefore, before drugs are prescribed, health care providers should inform their oldest old patients of the expected effects of treatment and possible side effects.

The frequency of PIM in this study was 27%. A study from the Middle East showed a PIM frequency of 52.5% [56]. A study from Taiwan revealed PIM frequencies varying between 24 and 73% among different hospitals [57]. In Europe, the frequency of PIM is around 20% [58], while it ranges from 14 to 27% in the United States [59]. Therefore, the frequency of PIM reported in this study was comparable to that observed in other countries, and lower than in previous studies in China. Indeed, among elderly ≥ 65 years of age, Li et al. reported a PIM frequency of 52% or 72%, depending upon the definition used [60], while Zhang et al. [61] showed PIM rates of 45–54% among hospitalized patients. These discrepancies could be due to the study population (in this study, oldest old ≥ 80 years of age), the definition of PIM, and period covered (here, pre-admission). The benzodiazepines were commonly used of PIMs In this study. The oldest old who took the benzodiazepines were more likely to be cognitive and psychomotor impairment compared with young people. So No-drug Therapy is advised firstly for the oldest old with insomnia.

The present study is not without limitations. The sample size was relatively small and from a single center. The selection of the patients will inevitably result in biases, affecting the generalizability of the results. Finally, no control group was included and the analyses were mostly descriptive, limited the internal and external validity of the results. Additional studies are still necessary to improve drug management in elderly patients.