Medicine use by the elderly in Goiania , Midwestern Brazil

OBJECTIVE: To analyze the pattern of use of medications use in aged people and associate it with socioeconomic aspects and with the self-rated health. METHODS: A population-based cross-sectional design study with 934 elderly people from Goiania, Midwestern Brazil, between December 2009 and April 2010. Data were collected through a questionnaire. The dependent variable was the number of medications consumed and the independent variables were sex, marital status, education, type of residence, age, income, and self-rated health. Drugs were classifi ed according to the Anatomical Therapeutic Chemical Classifi cation. The inappropriate drugs for the elderly were identifi ed according to the Beers-Fick criteria. The tests used were Chi-square and Fisher’s exact test, p was considered signifi cant when < 0.05. RESULTS: The elderly consumed 2,846 medicines (3.63 medications/person). The most commonly consumed were those which act in the cardiovascular system (38.6%). The prevalence of polypharmacy was 26.4% and selfmedication was 35.7%. The most used drugs for self-medication were analgesics (30.8%), 24.6% of the elderly consumed drug considered inappropriate. Women, widows, those aged 80 or over and with worse self-rated health were more likely to practiced more polypharmacy. Most self-medication was associated with lower levels of education and worse self-rated health. CONCLUSIONS: The pattern of drug use by the elderly was similar to that found in the elderly in other regions of Brazil. The number of drugs used, the prevalence of self-medication and practice of polypharmacy and inappropriate drug use were within the national average. DESCRIPTORS: Aged. Drug Utilization. Drug Therapy, Combination. Drugs of Continuous Use. Self Medication. Diagnostic Self Evaluation. Cross-Sectional Studies. Medicine use by the elderly Santos TRA et al Ageing populations are considered a global phenomenon and constitute one of the biggest public health challenges we face today.3 Brazil also fi nds itself in this situation. The 2010 Population Census showed that the elderly make up 12.0% of the population. In Goiás, the elderly constitute 9.3% of the total population and 9.6% in Goiânia, Midwestern Brazil.a Higher prevalence of chronic illness means that the elderly are frequent health care service users and possible society’s most medicated group. The elderly population contribute around 25.0% of total medicine sales in developed countries.2 Therapeutic proposals making use of various medicines simultaneously are inappropriate and may have serious, even fatal, consequences for the elderly due to alterations in the metabolism which occur with age.17 Failure to follow pharmacological treatment, adverse reactions, drug interactions, high prices and hospitalizations are the main consequences of practicing polypharmacy.7 Elderly bodies present alterations in physiological functions which should not be neglected. These alterations lead to different pharmacokinetics and greater sensitivity to both the therapeutic and adverse effects of medicines.19 Some medicines are considered unsuitable for use by the elderly due to reductions in their effi cacy or because the increased risk of adverse effects outweighs their benefi ts.9 Self-medication places the health of the elderly population at risk. This practice can increase risks related to prescribed medicines, delay correct diagnosis and mask illness.23 Knowledge of the elderly population’s medicine use and factors related to this is indispensable in order to be able to redefi ne public policy and improve the elderly’s quality of life and health. This study aims to analyze medicine use patterns in the elderly and their links to socio-economic aspects and self-rated health.

Ageing populations are considered a global phenomenon and constitute one of the biggest public health challenges we face today. 3Brazil also fi nds itself in this situation.The 2010 Population Census showed that the elderly make up 12.0% of the population.In Goiás, the elderly constitute 9.3% of the total population and 9.6% in Goiânia, Midwestern Brazil.a Higher prevalence of chronic illness means that the elderly are frequent health care service users and possible society's most medicated group.The elderly population contribute around 25.0% of total medicine sales in developed countries. 2erapeutic proposals making use of various medicines simultaneously are inappropriate and may have serious, even fatal, consequences for the elderly due to alterations in the metabolism which occur with age. 17Failure to follow pharmacological treatment, adverse reactions, drug interactions, high prices and hospitalizations are the main consequences of practicing polypharmacy. 7derly bodies present alterations in physiological functions which should not be neglected.These alterations lead to different pharmacokinetics and greater sensitivity to both the therapeutic and adverse effects of medicines. 19Some medicines are considered unsuitable for use by the elderly due to reductions in their effi cacy or because the increased risk of adverse effects outweighs their benefi ts. 9 Self-medication places the health of the elderly population at risk.This practice can increase risks related to prescribed medicines, delay correct diagnosis and mask illness. 23owledge of the elderly population's medicine use and factors related to this is indispensable in order to be able to redefi ne public policy and improve the elderly's quality of life and health.This study aims to analyze medicine use patterns in the elderly and their links to socio-economic aspects and self-rated health.

METHODS
This is a descriptive, transversal population-based study of 934 elderly people from Goiânia, Midwestern Brazil, between December 2009 and April 2010.
Sample size was calculated based on the number of elderly people in the municipality in 2007 (7.0% of 1,249,645 inhabitants) and 30.0%expected frequency for the specific objectives of the epidemiological survey, with 955 confidence interval, 5% level of signifi cance, absolute precision of 5% and 1.8 design effect of cluster sampling (DEF).To cover possible

INTRODUCTION
losses, an extra 11% was added, giving a sample of 934 elderly people.Individuals aged ≥ 60 and who resided more than four days a week in the homes where the interviews took place were considered eligible to take part.Those who were not at home at the time the visit took place were excluded.The interviewers had been previously trained by REVISI staff.
A pilot study with 50 elderly people was carried out in order to assess the instrument to be used in the research and to identify the mean number of elderly people in each census sector (CS) of the municipality.
Each CS had a mean of 17 elderly people.Thus, 56 CS would be necessary to reach the sample size.The CSs were randomly selected from the 912 CS which make up urban Goiânia, using an electronic randomizing system.Draws took place to defi ne the blocks and corners in which data collection should begin for the selected CS.
The fi rst residence from the corner selected, excluding non-residential properties, was visited.If no elderly person lived there, the interviewer moved on to the next residence until an elderly resident was found, after which the systematic search began again.If two consecutive houses visited had elderly residents the second house was not included.And so on, until the estimated total of elderly people for the sample was reached.
Those elderly people eligible to take part were invited to participate in the research and respond to a questionnaire.This questionnaire was composed of 12 sections and delivered to central offi ce for quality control of the data after being completed.
The dependent variable was the number of medicines used and the independent variables were gender, marital status, schooling, type of housing, age and self-rated health.
The following questions were used: "Which medicines do you use regularly?"and "Was the medicine prescribed/recommended by the doctor, a neighbor, pharmacy employee, yourself, family member or from an old prescription?"The elderly person was asked to show what medicines they took, with their respective prescriptions, and the names were copied directly from the labels.Prescriptions were considered old if they were dated more than three months before the interview.
The information from the questionnaires were used to verify associations between polypharmacy and self-medicating with socio-economic conditions and self-rated health and between polypharmacy and a Instituto Brasileiro de Geografi a e Estatística.Censo 2010.Brasília; 2011 [cited 2011 Jun 20].Available from: http://www.censo2010.ibge.gov.brb World Health Organization.Anatomical therapeutic chemical -ATC classifi cation index with defi ned daily doses-DDD's.Olso; 2000.self-medicating.The number of medicines was counted based on the elderly's reports and polypharmacy was defi ned as using fi ve or more medicines. 13Self-rated health was assessed using the following question: "In general, would you say your health was very good, good, regular, bad or very bad?" Information on gender, age, schooling, marital status, family income and type of housing was collected for the socio-economic profi le.
The main active ingredient of each medicine was identifi ed and classifi ed according to the Anatomical Therapeutic and Chemical Classification -ATC b into 14 major groups, corresponding to an anatomic number.Each group was divided into subgroups, corresponding to therapeutic groups.These two fi rst groups (anatomic group and therapeutic subgroup) were those taken into account for this study.This classifi cation does not include natural products classifi ed as phytotherapeutic.
The Beers-Fick 9 criterion was used to identify prevalence of use of medicines unsuitable for the elderly.The main active ingredient of each of the medicines used by the elderly in the study was checked to see whether it was in the list of this criterion.Numerical variables were explored using descriptive measures of centrality and dispersion and categorical variables were explored using simple and relative frequencies.To investigate the association between qualitative variables the Chi-square (X 2 ) test was used n ≥ 5 and Fisher's exact test when n < 5; prevalence ratio (PR) and their respective confi dence intervals (IC).The signifi cance level for all tests was set at p < 0.05.

RESULTS
Of the elderly people studied, 83.8% fully completed the questionnaire.Of these, 65.0% were female, 45.7% were aged between 60 and 69 and the mean age was 71.9 (minimum 60 and maximum 96 years old).Around 50.0% had started or completed Primary Education, 49.5% were married and 32.0% widowed and 80.0% of the elderly in this study lived in their own home.They predominantly (46.5%) rated their own health as 'regular'.
In this study, 2,864 medicines were used by the elderly (a mean of 3.63 medicines per person) and they had taken at least one medicine on the day of the interview.The maximum number of medicines was 19.
Schooling, age and self-rated health proved to be associated with self-medicating (p < 0.05).This practice was less frequent among individuals with higher levels of education (RP = 0.52), those aged > 80 (RP = 0.74) and those who rated their own health as very good (RP = 0.43).There was no link between polypharmacy and self-medicating (p > 0.05) (Table 5).

DISCUSSION
The results of this study show that the mean number of medicines used by the elderly in the municipality of Goiânia was higher than those observed in other state capitals in Brazil such as Porto Alegre, Southern Brazil and Fortaleza, Northeastern Brazil 6,10 and lower than those observed in Belo Horizonte, Southeastern Brazil. 21These inequalities in the numbers of medicines used may be explained by differences in the services provided to the population and the type of health care model used in each region. 14The continued growth in the number of medicines used among the elderly may be explained by the increase in prevalence of chronic illness in this age group, as well as by a health care model which views medicine as the main form of intervention.However, the implications of this use need to be measured and assessed as to their risks benefi ts. 7ere was a high prevalence of polypharmacy, as more than ¼ of the elderly people were exposed to this practice.This prevalence was greater than that observed in Belo Horizonte and similar to that found in Porto Alegre. 10,14National studies indicate polypharmacy prevalence of between 11.0% and 40.6%. 14,18lypharmacy is often necessary, as many elderly people suffer from a variety of illnesses and symptoms which require the use of various medicines to guarantee the best possible quality of life.Polypharmacy with medications which have not been prescribed does not necessarily mean that these medicines are being used incorrectly. 7However, there are high rates of polypharmacy and the use of multiple medications increases the risk of adverse reactions and drug interactions. 7,14A more critical and systematic approach for those elderly who really need multiple medications is needed.
The most elderly are those who most frequently practice polypharmacy, as has been seen in other studies in Brazil. 14,21This correlation between polypharmacy and increased age may be related to higher use of health care services by those of most advanced age. 13 was women who most frequently practiced polypharmacy, similar to other national epidemiological studies. 21,22Greater use of medicines by elderly women may be connected with questions such as women living longer than men and thus spend longer living with chronic illness, the greater attention women pay to their own health problems and reported higher demand for health care. 13e number of elderly people who self-medicate is similar to that found in other Brazilian studies.More than 30.0% of the elderly in the north east self-medicated and a prevalence of 46.0% was observed in Minas Gerais. 6,15A revision of studies on self-medication among the elderly shows rates between 12.0% and 44.0%. 1,16A large part of the population, including the elderly, self-medicate to treat minor and major symptoms.Reports indicate easy access to medicines as an important factor in this practice. 15e most commonly self-medicated medicines are analgesics, as has been observed in other Brazilian studies. 4,14Self-medicating with analgesics is typically common among the elderly, as their use is related to treating pain and infl ammation, common symptoms at this stage.
Medication for high-blood pressure occupies the second place in the most commonly self-medicated drugs.This is worrying, even when they have been previously prescribed.Hypertension is a chronic disease with major health problems and using medications to control it requires regular monitoring.The major cause of mortality among the elderly in Brazil is the stroke, probably due to a lack of systematic control of risk factors. 12derly people with lower levels of education are those who most frequently self-medicate.In a study carried out in the Northeastern Brazil, the least well off elderly were those who practiced the most self-medication. 6his may be explained by diffi culties accessing health care services and prior awareness raising on the risks self-medicating cause.Elderly people aged 80 and over were those who were less likely to self-medicate, possible due to higher levels of using health care services where the patient is better cared for. 13 link was found between polypharmacy and selfmedication.The high prevalence of polypharmacy was due to prescriptions and not to self-medicating.This shows the importance of raising doctors' awareness about medicalization, especially among the elderly. 7er half the elderly rated their own health as regular, bad or very bad.This fi nding was better than that observed in Bambuí 11 and similar to that found in Lima-Costa et al, 12 using data from the 2003 PNAD (Pesquisa Nacional por Amostragem Domiciliar -National Research for Sample of Domicile), in which 56.4% rated their own health as regular, bad or very bad.Classifying health as very bad was associated with polypharmacy, similar to the fi ndings of other pharmacoepidemiological studies. 11,14Patients who rated their health as very bad sought solutions to their health problems in medicines, whether prescribed or self-medicated.
The prevalence of use of medicines unsuitable for the elderly is within the range found in other studies in Brazil (15.4% to 41.0%). 8,18This high prevalence is largely due to prescribed medicines.In a study carried out in Rio de Janeiro, Southeastern Brazil, 17.0% of medicines used were unsuitable, of which 90.0% had been prescribed. 22his refl ects the lack of knowledge of medicines which are inappropriate for the elderly on the part of the doctors, something which could have serious clinical and economic consequences for the health care system. 9e most commonly used unsuitable medicines in this study were benzodiazepines and anti-depressants.Despite the serious consequences using these medicines may have for the elderly, this fi nding was similar to that of Castellar et al. 5 Benzodiazepines have a longer half-life in the elderly and consequent prolonged sedation with the risk of falls and fractures.The same occurs with anti-depressants, with which there is a strong possibility of anticholinergic effects (diffi culty breathing, blurred vision, increased heart rate, and decreased blood pressure), orthostatic hypotension and stimulation of the central nervous system. 9ten, the pharmacological profi le of these medicines and their possible consequences when used by the elderly is not known to the doctor. 7This situation is not improving.The very act of visiting a health center increases the chance of using unsuitable medicines, as the SUS (Brazilian Unifi ed Health System) does not have a list of medicines more suitable for elderly.Obreli Neto & Cuman 20 (2010) verifi ed a signifi cant reduction in use of unsuitable medicines by patients monitored by a multi-professional program.
The most commonly used groups of medicines were similar to those in the national and international literature. 6,10,14,16,21Medicines for the cardiovascular system were the most commonly used, explained by the high prevalence of cardiovascular disease among the elderly population. 12In the therapeutic group, medicines for high blood pressure were those most commonly used, which corresponds to the high levels of hypertension in Brazil. 11though the results of this study are important, some limitations should be considered.The questionnaire involved various aspects of the elderly's health and was not specifi c to medicine use.Such observations, however, do not compromise the importance of the study.Subsequent investigations should evaluate medicine use among the elderly in a broader way, looking at adhesion, dosage, posology and method of administration.
The pattern of medicine use among the elderly in Goiânia was similar to that found in the elderly in other regions of Brazil.The number of medicines used, the prevalence of polypharmacy and self-medication and use of unsuitable medicines were within national average.The most commonly used groups of medicines corresponded to the most common illnesses in the elderly population and concur with various other studies in Brazil.Women, widowers, those aged 80 and over and those who rated their own health as very bad were those who most commonly practiced polypharmacy; self-medication was associated with lower levels of education and worse self-rated health.
Health care professionals' contributions are necessary in order to optimize the rational use of medicines for the elderly and reduce as far as possible the complications linked to their use.The results of this study may serve as an alert to health care managers, with the aim of adapting the health care network to the real needs of today's elderly as well as preparing for the new contingents of elderly which increase every year.
The data were inputted into the Excel program.Statistical analysis was carried out using Statistical Package for the Social Sciences (SPSS) version 15.0 for Windows and EpiInfo 6.04.The study was approved by the Ethical Research Committee of the Universidade Federal de Goiás (Protocol nº 050/2009).

Table 1 .
Distribution of medicines unsuitable for use by the elderly, according to Beers-Fick a criterion based on the pharmacological group and possible consequence of use.Goiânia, Midwestern Brazil, 2010.

Table 2 .
Distribution of medicines used by the elderly according to ATC anatomical and therapeutic classifi cation.a Goiânia, Midwestern Brazil, 2010.

Table 2 .
Continuation a Anatomical Therapeutic and Chemical Classifi cation

Table 3 .
Frequency of origin of recommendation/prescription of medicines used for self-medication.Goiânia, Midwestern Brazil, 2010.
a Chi-square test and Fisher exact test ni: not included