Pharmacotherapy follow-up in elderly in a public outpatient clinic in Brasilia , Brazil Mirna

The aging process is followed by an increase in the number of chronic diseases and, thus, the use of multiple medicines and prevalence of drug-related problems (DRP). This paper aims to identify the impacts of Pharmacotherapy follow-up (PF) in elderly with chronic non-communicable diseases while in polypharmacy using, in the public health service of Brazil. A prospective cohort study with 45 elderly of both sexes was done. Socioeconomic and pharmacotherapy data was collected at the first pharmaceutical appointment. According to the pharmacotherapeutic needs of each case were applied interventions for four appointments as blood pressure and blood glucose monitoring, schedule control medicine use and prescription changes by the letter to prescriber. Some variables as blood pressure, blood glucose, body weight, adverse drug reactions, negative outcomes associated with medication, drug-drug interactions and adherence were compared after 3, 6 and 12 months of follow-up during baseline in order to evaluate the impact of the study. Sample included elderly with a mean age of 71 years old, mainly low income and low education women. After 12 months of PF, pharmacotherapy adherence has shown an improvement (21.3 vs 52.4%), as well as a decrease in the number of adverse drug reactions (-1.61), drug-drug interactions (-1.17) and negative outcomes associated with medication (-2.49). Besides, there was better control of systolic blood pressure (-5.89 mmHg), body mass index (2.01) and blood glucose (-12.51 mg/dl) besides plasmatic levels of lipoproteins, vitamin D and thyroid hormones. Results show evidence of PF importance as a strategy to improve adherence to pharmacological treatment, reduce drug-drug related problems and promote better quality of pharmacotherapy. Hence, better health for medication users.


INTRODUCTION
As age progresses, so do physical, motor and cognitive limitations.In connection to multimorbidity, to pharmacokinetics and pharmacodynamics characteristics and polypharmacy, those limitations make the elderly a vulnerable group to developing and worsening drugrelated problems (DRP).In addition, negative outcomes associated with medication (NOM) also appear (Chua et al., 2012;Silva et al., 2013).The main DRP in these patients include lack of information about the disease and pharmacological treatment, lack of adherence to treatment, plus incorrect administration and problems related to drug safety (Chua et al., 2012;Lyra Jr et al., 2007;Silva et al., 2013).This situation gets worse due to lack of individualized counseling and to worse socioeconomic conditions (Cho and Kim, 2014;Chua et al., 2012;Sirey et al., 2013).
Many studies have shown that such problems are reduced when multidisciplinary monitoring is introduced to the patient in his home, institution, hospital and outpatient clinic context.Integrating a clinical pharmacist to the staff has shown to reduce the number of prescribed medications, to improve the adhering profile to the therapeutic scheme, to reduce the number of admissions due to NOM and adverse drug reactions (ADR).Moreover, it has shown to decrease health assistance costs, fostering improvement of the patient's quality of life altogether (Aguiar et al., 2012;Balisa-Rocha et al., 2014;Castro et al., 2006;Chisholm-Burns et al., 2010;Correr et al., 2011;Hayward et al., 2015;Lyra Jr et al., 2008;Mansoor et al., 2014;Morgado et al., 2011;Plaster et al., 2012;Scotti et al., 2013;Souza et al., 2007;Stewart et al., 2014).
Studies on pharmacotherapy follow-up show a reduction of approximate 80% of DRP numbers and significant adherence improvement in various groups of patients (Chua et al., 2012;Mansoor et al., 2014;Lee et al., 2006;Silva et al., 2013).Likewise, studies have shown a decrease in using emergency services, followed by an increased search for primary services, thus, an increase in direct and indirect economic benefits.For example, reduction in admissions and disease complications costs (Borges et al., 2011;Chisholm-Burns et al., 2010;Hayward et al., 2015;Scotti et al., 2013).Pharmacotherapy follow-up in hypertensive and diabetic patients, for example, is related to a better control of blood pressure and glucose levels with decrease of cardiovascular risk (Aguiar et al., 2012;Castro et al., 2006;Correr et al., 2011;Hayward et al., 2015;Lyra Jr et al., 2008;Morgado et al., 2011;Plaster et al., 2012;Scotti et al., 2013;Souza et al., 2007;Stewart et al., 2014).Therefore, taking into account the importance of this model of professional practice with benefits in the health of the medicines user´s and public health system, this paper aimed to evaluate the impacts of pharmacotherapy follow-up in the health and pharmacotherapy quality of elderly cared for in a public outpatient clinic of the Oliveira and Novaes 819 Federal District, Brazil.

METHODOLOGY Study design and localion
A prospective cohort study was done.The patients were recruited from June to August, 2013 in a public ambulatory.This service is linked to Health Department, which is a reference in Geriatrics in Brasilia, capital of Brazil.The first appointments were all realized in September, 2013 and follow-up was given from October, 2013 to 2014.

Sample and inclusion and exclusion criteria
The sample was chosen for convenience including 45 patients who had appointments in the Division of Geriatric Services and that met the inclusion criteria.There were included in this study the seniors over 60 years old, of both genders, with non-communicable chronic diseases as systemic arterial hypertension, diabetes mellitus, osteoporosis and hypothyroidism diagnosed by a physician and pharmacological treatment.They were found by the medical or nursing staff who either had problems with adherence to the pharmacological treatment or made use of polypharmacy; they were taking at least one continuous medicine and were able to be interviewed and oriented by the researchers.Those under 60 years of age, with infectious transmitted diseases, dementia, visually or hearing impaired were excluded, as well as those with lack of discernment who did not make continuous use of medication.In Brazil, more than 60-year-old person is considered elderly (Brazil, 2003).After the beginning of the study, patients who failed attending at least one consultation or failed performing the lab tests were excluded from the follow-up.

Data collection and outcomes measures
Data was collected between the years of 2013 and 2014 through a semi structured questionnaire, which was adapted to The Dader Method for Pharmacotherapy follow-up (2003) (Machuca et al., 2003).The clinical pharmacist applied the previously scheduled questionnaire during appointments in the location of the study.Each patient had to be seen five times by the same clinical pharmacist.In service there are no clinical pharmacists.All pharmacists only develop functions related to supply, logistics and dispensing drugs without pharmaceutical orientation.During the first appointment, the following were evaluated: state of health, life style and current pharmacological treatment.Medications in use, current prescriptions and results of last exams were also evaluated.In cases when the patients did not have updated lab exams (last six months), the pharmacist asked for the following ones: hemogram, glucose, post-prandial glucose, glycated hemoglobin (A1C),lipid profile, urea, creatinine, uric acid, glutamic oxaloacetic transaminase (GOT), glutamic pyruvic transaminase (GPT), creatine phosphokinase (CPK), electrolytes (Na, K, Mg, Cl and Ca), TSH, free T4, vitamin D3 25-hydroxy, vitamin B12 and parathyroid hormone (PTH).The tests were done in the study lab partner.The reference values for the lab tests were: hematocrit (33 to 47.8%), glucose (70 to 99 mg/dl), post-prandial glucose (less than 140 mg/dl), A1C (less than 7%), total cholesterol (less than 200 mg/dl), LDL cholesterol (less than 100 mg/dl), HDL cholesterol (40 to 60 mg/dl), VLDL cholesterol (less than 30 mg/dl), urea (13 to 43 mg/dl), seric creatinine (for men 0.7 to 1.2 mg/dl; for women 0.53 to 1.0 mg/dl), uric acid (glutamic for men 3.5 to 7.2 mg/dl; for women 2.6 to 6.0 mg/dl), GOT (for men 13 to 40 U/L; for women 13 to 35 U/L), GPT (10 to 49 U/L), CPK (for men 31 to 294 U/L; for women 33 to 211 U/L), Na (132 to 146 mEq/L), K (3.5 to 5.5 mEq/L), Mg (1.3 to 2.7 mg/dl), Cl (99 to 109 mEq/L), Ca (8.3 to 10.6 mg/dl), TSH (0.35 to 5.5 mcUI/ml), free T4 (0.7 to 1.8 ng/dl), vitamin D3 25hydroxy (more than 30 ng/ml), vitamin B12 (180 to 900 pg/ml) and PTH (12 to 72 pg/ml).
After determining the situation and identifying DRP and NOM, pharmacotherapy care plans were defined to each patient in order to promote a change in his health and the quality of pharmacotherapy.In some specific cases, to plan interventions, an analysis of the patient's medical records was necessary.DRP and NOM concepts complied with the Third Consensus of Granada (GIAF-UGR, 2007).
In the next four appointments (1, 3, 6 and 12 months after the first one), necessary pharmaceutical interventions were applied and the main indicators of state of health and pharmacotherapy quality were reevaluated.That included adherence to pharmacotherapy; number of complaints, drug-drug interactions (DDI), adverse drug reactions (ADR), NOM; body weight, blood pressure, glucose levels, lipid profile, vitamin D and thyroid hormones control.
In order to evaluate adherence to pharmacotherapy, the patient was considered to have adhered when he did not present any problems related to medication, such as times, posology, types of administration and dosage.For that, some indicators validated by the Morisky-Green tests and the brief medication questionnaire (BMQ) were used (Ben et al., 2012).Complaints were identified through the patient's medical background, while the pharmacist performed the anamnesis and asked about possible complaints (pain or discomfort) that affected different parts of the body.The DDI were initially evaluated by Prescription Checker application version 1.3.1/07-2014available on iOS Apple® system developed by Chaillet (2014) and confirmed by specific literature (Bachmann, 2006).

Data analysis
At the end of the fourth appointment, the variables of the baseline were compared to the final state (after interventions) aiming to get to know the influence of pharmacotherapy follow-up in the group being studied.For ethical reasons, pharmaceutical interventions were applied to all research participants.The patient himself was used as control, and the results of each appointment were compared to the initial data (baseline).Data was saved in Microsoft Excel® spreadsheets and analyzed by Statistical Analysis System (SAS) version 9.3 by a statistician.Quantitative variables were analysed using Chisquare and Student´s t tests taking into consideration significant results to p ≤ 0.05.Initially, mixed effects models of regression, with linear or quadratic tendency, were applied to test the intervention effect throughout the follow-up period.When the p-value for linear or quadratic tendency was lower than 0.005, pre-specified comparisons at 3, 6 and 12 months were tested in comparison to baseline.That was done with the use of mixed effects models of analysis of variance for repetitive measures adjusted for baseline values.The main focus of the analysis was the change at 3, 6 and 12 months in comparison to baseline.The Bonferroni correction method was used to adjust the three prespecified comparisons.In order to compare adherence, the MacNemar Test was applied to compare percentage after 6 and 12 months of follow-up to baseline.To evaluate the correlation between the quantitative variables, the Pearson linear correlation coefficient and the Fischer exact test were used in order to to evaluate the link between the categorized variables.For all tests, a significant level of 5% was considered.

Ethical aspects
The

Paper review
The paper was reviewed according to STROBE Statement-Checklist to cohort studies.The keywords are indexed to MeSH terms.

RESULTS
The initial sample had 105 elderly of both sexes who met the study requirements.Of those, 57.1% (n = 60) abandoned the study due to quitting, death, moving out of city, changing health services, not going to scheduled appointments and/ or taking the prescribed tests (Table 1).The group that remained in the study until the end (n = 45) was with a mean age of 71 years old (standard deviation (SD) ± 5.2), mainly 70 to 74 years old and 88.9% women.The social demographic profile of the group shows that most of the elderly came from the northeast region of Brazil (44.4%), widowhood (42.2%), with an average family income of 3.2 minimum salaries (1 MS = R$ 724.00), with 4 to 7 years of formal education (33.3%) (Table 2).
There was no significant difference in the number of medications taken by men and women (p = 1.0000).However, statistical analysis showed that there was a positive and significant correlation between the number of taken medications, the number of health problems (r = 0.53; IC 95%: 0.28 to 0.71; p < 0.0001) and the number of self-related complaints (r = 0.30; IC 95%: 0.01 and 0.54; p = 0.0394).In relation to the pharmacotherapeutic profile, the study showed that all elderly (100%) took their medications by themselves.However, only 82.2% (n = 37) presented problems related to pharmacotherapy adherence mainly because of drug reactions (73.3%), life habits (55.5%), health conditions (44.4%) and lack of information about the pharmacological treatment (40.0%).71.1% (n = 32) of the group reported self-medication practices.The statistical analysis did not show correlation between adherence and age (p = 0.2858), gender (p = 1.0000), income (p = 0.2806), family structure (p = 1.0000), number of complaints (p = 0.7238), ADR (p = 1.0000) and NOM.Nonetheless, during baseline, the percentage of low education patients who adhered to treatment was significantly lower (p = 0.0175 -Fisher exact test).They identified 177 ADRs (an average of 3.9 per patient) in the sample and the most common were    4).
The number of ADR and NOM showed a positive and significant correlation with the number of complaints (r =  the number of DDI (r = 0.76; CI 95%: 0.59 to 0.86; p < 0.0001).
Besides the interventions, the following were applied: monitoring other laboratory parameters (thyroid hormones, urea, creatinine, creatinine clearance, GOT, GPT and others), referring to medical specialities (psychiatry, cardiology, dermathology and mainly ophthalmology), referring to a nutritionist or other health professional (such as a psychologist), giving information on access to some prescribed medicines and pharmaceutical prescription.Pharmaceutical interventions are shown in Table 5.It was observed that 61.5% (n = 247) of the applied interventions were accepted during pharmacotherapy follow-up.Most of the pharmaceutical interventions can be carried out without the medical approval except for changes in the number or types of the prescribed medicines.The acceptance was evaluated by physicians and/or patients.
After four pharmaceutical appointments, there was evidence of a significant change in the health and pharmacotherapeutic profile of the sample.Notice that these changes were possible thanks to all the applied pharmaceutical interventions.It was demonstrated improvement in the adherence profile to pharmacological treatment.At the end of the 12-month period, the percentage of patients adherent to treatment (52.4%) was significantly higher than the percentage of patients with adherence in baseline (21.3%) (p = 0.0008).When adherence was evaluated after 6 months of follow-up, there was evidence of an increase in the number of elderly who adhered (53.5%) (p = 0.0005) (Table 6).However, the adherence percentage did not change between the 6 th and the 12th months (p = 0.5637).Patients followed in the period of 12 months showed a very significant decrease in the number of complaints, ADR, NOM and DDI (Table 5).The average number of complaints decreased significantly throughout the followup period, and after 12 months there was an average decrease of 2.5 complaints (26% change compared to baseline).The average number of ADR decreased an average of 1.61 (43% change compared to baseline), NOM decreased an average of 2.49 (50% of change compared to baseline) while decrease was of 1.17 (26% of change compared to baseline) in the number of DDI (Table 7).
In relation to the laboratory parameters, at the end of the study, the patients showed a decrease in fasting glucose serum levels, total cholesterol (TC), LDL, triglycerides (TG), free T4, TSH and vitamin D. In addition, there was body weight loss and, therefore, body mass index (BMI) and systolic blood pressure (SBP).
There was no significant difference in the remaining parameters such as diastolic blood pressure (DBP) (Table 8).

DISCUSSION
The benefits of Pharmacotherapy follow-up (PF) evidenced by several studies include better control of chronic diseases such as hypertension, diabetes, dyslipdemia, osteoporosis, ashma, AIDS, depression, Parkinson, Alzheimer, among others.Moreover, it provides economic benefits (Chua et al., 2012;Morgado et al., 2011;Silva et al., 2013;Scotti et al., 2013;Souza et al., 2007;Stewart et al., 2014).Most part of the PF studies involves elderly patients due to pharmacotherapy profile characteristics, with increased predominance of DRP and NOM in group.Thus, it is believed that the results of this study might contribute to better understanding DRP in the elderly population as well as understanding the benefits of PF.Although this study is limited to a specific geographical region and the small sample size, the social economic characteristics and the pharmacotherapy are common to the the elderly population in general.Thus, it allows for important conclusions on the impacts of this new model of pharmaceutical practice.In this study, a higher number of women were already expected due to the demographic characteristics of the Brazilian population.According to the last demographic census, women make the majority in all age and senior groups; they comprise 55.5% of the population (IBGE, 2013).In a study developed by Silva et al. (2013), women were also the majority and the sample also included low income and education, a limiting factor to adherence to pharmacotherapy (Silva et al., 2013).Similar conclusions were reached in the work of Cho and Kim (2014) and Plaster et al. (2012) (Cho and Kim, 2014;Plaster et al., 2012).
The average consumption of drugs showed in this study (eight medicines/patient) is similar to Silva´s study done (Silva et al., 2013).The high level of drugs consumption in the studied sample might be explained by the number of health problems (an average of four per patient) and by the old age group (a mean age of 71 years old).In this study, an average of eight medicines per patient makes the complex therapeutic regimen, considering the age group.Polypharmacy in elderly is associated with worse health conditions and significant economic impacts.This study showed a correlation between the number of complaints and health problems, which might reflect health conditions, and the number of medicines taken.That corroborates Kadam (2011) findings, which showed a positive association between polypharmacy and worse conditions in the physical and psychological health of British individuals over 50 years old (Kadam, 2011).Huri and Ling (2013) also showed a relation between polypharmacy and DRP in patients with diabetes type 2 and dyslipidemia.According to the authors, there is a strong relation between the lack of control of lipoprotein levels, kidney diseases and DRP (Huri and Ling, 2013).
DRP found in this study, such as self-medication, drug interactions, lack of information about the treatment and lack of adherence were also found in other studies involving elderly patients (Cho and Kim, 2014;Morgado et al., 2011;Pittman et al., 2010).Adherence to the pharmacological treatment is one of the main concerns when dealing with the elderly's care.In this study, the high percentage of patients with difficulty to adhering to the pharmacological treatment is due to many factors: old age, social economic level (low education and income levels), health conditions (prevalence of diseases and complaints), difficulty to access health services, complexity of the therapeautic regimen and prevalence of adverse reactions similar to those found by Cho and Kim (2014), even though the statistical analysis did not show a correlation between DRP and NOM and adherence, probably due to the sample size.e p-value compared to the follow-up to baseline calculated using a mixed model of analysis of variance for repeated measures and effects were shown when the p value was less than 0.05 for the effect of the intervention over time.Bonferroni correction was applied to adjust the pre-specified comparisons.
The complexity of a patient's therapeutical scheme represents an important factor when evaluating adherence to pharmacotherapy.An estimate of 40 to 75% of elderly do not follow their therapeutic regimens correctly due to an increase in physical and motor limitations, cognitive deficit, not being able to understandinstructions, something common to age, moreover due to lack of communication, not having a caretaker and even due to the complexity of the treatment (Acurcio et al., 2009;Sirey et al., 2013).Vries et al. (2014) showed in their study about diabetes type 2 a relation between the complexity of the therapeutical regimen and the lack of adherence, concluding that simpler regimens shoud be given priority as a strategy to improve adherence (Vries et al., 2014).Laliberté et al. (2012), for example, showed that adherence is 26% higher in patients making use of simpler therapeutic regimens (once a day) than of those more complex (three times/day) (Laliberté et al., 2012).
Lack of adherence is related to difficulty with controling blood pressure, levels of plasma lipids, glucose and other clinical and lab parameters.This problem is related to worsening and developing complications (Devold et al., 2012;Hiligsmann et al., 2012;Peyrot et al., 2012;Pittman et al., 2010;Simpson et al., 2011).Devold et al. (2012) in a study with many bi-phosphonates, showed that patients have a tendency to develop adherence problems during chronic treatments throughout the period.In addition, they have even more difficulty to follow the prescription during the first year of treatment (Devold et al., 2012).The authors concluded that adverse reactions to medication, difficulty to administer, problems with forgetting and lack of a doctor's or nurse's follow-up, age, family income, education level and marital status e p-value compared to the follow-up to baseline calculated using a mixed model of analysis of variance for repeated measures and effects were shown when the p value was less than 0.05 for the effect of the intervention over time.Bonferroni correction was applied to adjust the pre-specified comparisons.
Among the elderly being studied, the high level of DRP (3.9 per patient) and NOM (5.0 per patient) might be explained by the fact that 100% of the sample was using polypharmacy.Besides that, the group's old age and itsvisual, hearing, motor and cognitive limitations in connection to many complaints might contribute to increasing these figures.The highest number of NOM of the safety group is consistent with data related to complaints and DRP.Therefore, there is a statistical correlation between those variables.In Silva et al (2013) and Plaster et al (2012) the main problems were also related to the safety of the medicines (Silva et al., 2013;Plaster et al., 2012).
The results of NOM of the necessity and effectiveness groups deserve some attention because they might be reflecting the quality of the services rendered to the group, indicating the need to invest on hiring more health professionals as well as on training them.Such data reinforces how important the multiprofessional follow-up is when it comes to health assistance, including the clinical pharmacist in the staff and implementing a continuous flow of services to these patients through regular monitoring.It is important to highlight that the process of identifying DRP and NOM is influenced by the professional's own experience and the information given by patients themselves in the medical records.In health services where this information is not duly systemized, these numbers might be under or over estimated.In the place where this study took place, the medical records were 100% manual and the information was not always duly organized.Besides that, in this study, variables were identified by observation and specific tools or tests were not used for that purpose, which might have influenced the values.
The incidence of DDI also strongly influences the data of DRP and NOM and represents a predisponent factor for worse health conditions, bad pharmacotherapy and health assistance quality.The high number of DDI with moderade seriousness of risk C corroborates the findings of Brazilian´s study (Venturini et al., 2011).The risk index indicates which clinical proceedings should be adopted when facing the DDI while the seriousness indicates its possible consequences.Type C interactions with moderate seriousness in general require therapy monitoring and the benefits of association generally overcome the risks which are low and tolerated (Bachmann, 2006).
In this study, identifying DRP allowed for later applying different types of pharmaceutical interventions according to the need of each clinical case.The high number of DRP justifies the number of applied interventions (8.6 per patient).The most easily applied interventions were those that allowed direct contact between the pharmacist and the patient, such as monitoring body weight and organizing times for drugs administration using standard tables.Interventions that might require a third person, such as a doctor, were the most difficult to be implemented, especially when this professional was from a different health service.The high percentage of accepted interventions is due mainly to the cooperation of the medical staff in the location chosen for the study.Interventions that required a change of routines or life habits were the most difficult to be implemented and the least accepted by the patients therefore with the least effectiveness.Such results show that beliefs and habits in old age groups represent a predisponent factor to lack of adherence and worse health conditions (Gujral et al., 2014;Sirey et al., 2013).
Study conducted by Jódar-Sánchez et al. ( 2014) showed high acceptance rate (88.7%) of pharmacist reccomendations by institutionalized elderly pharmacotherapy follow-up for 12 months (Jódar-Sánchez et al., 2014).Gujral et al. (2014), in regards to adherence, when evaluating the impact of interventions from community pharmacists with patients with coronary diseases in regards to adherence, concluded that beliefs, as well as being monitored by the same or different professionals, influence behavior and adherence to pharmacological treatment in a very significant manner (Gujral et al., 2014).
It was also noticed that interventions that required some type of financial expense, such as exams or medical appointments not covered by that health service, were less accepted or less effective, which shows how important purchasing power is to guarantee access to better health conditions.It is relevant to note that in the sample, only 13.3% (n = 6) had health insurance with an average of family income of R$ 2,316.00 and low education level, factors that influence a lot this result.
Many studies have been published in the last decade showing the benefits of these interventions.For example, a better control of blood pressure, glucose, levels of plasma lipoproteins and vitamin D.Moreover, a delay in the development of some diseases and a lower risk of complications (Chisholm-Burns et al., 2010;Hiligsmann et al., 2012;Laliberté et al., 2012;Mansoor et al., 2014;Peyrot et al., 2012;Scotti et al., 2013;Silva et al., 2013;Stewart et al., 2014).Lowrie et al (2014) also showed that patients with chronic heart failure that had been cared for by a pharmacist acknowledged the importance of this professional to promoting information and adherence to the pharmacological treatment (Lowrie et al., 2014).The findings of this study show a significant improvement in adherence to the pharmacological treatment after 6 and 12 months of follow-up.However, between the 6th and the 12th months, this variable improvement was not identified.A study done by Mansoor et al. (2014) involving community pharmacists in Australia showed that 98% of those professionals acknowledge the importance of their professional role in promoting adherence to the pharmacotherapy although the use of strategies to improve adherence in the work routine is not that frequent (Mansoor et al., 2014).Reviewing the literature developed by Morgado et al. (2011), it was possible to identify many studies that showed an improvement in the adherence and in the blood pressure control in different groups of patients (Morgado et al., 2011).Stewart et al. (2014) also showed in their study Hypertension Adherence Program in Pharmacy (HAPPY) developed in Australia, a decrease in systolic blood pressure and an increase in adherence to the antihypertensive treatment after 6 months of follow-up in a group of 395 hypertensive individuals (Stewart et al., 2014).The work developed by Simpson et al. (2011) showed a better control of blood pressure in patients with diabetes type 2 with a decrease in complications after applying pharmaceutical care in primary care (Simpson et al., 2011).Fikri-Benbrahim and associates also showed an increase in the percentage of adherence after six months of follow-up in the provinces of Jaen and Granada (Spain) (Fikri-Benbrahim et al., 2013).Scotti et al. (2013) showed that controling adherence to antihypertensive treatment reduces cardiovascular risk with financial impacts (Plaster et al., 2012;Scotti et al., 2013).
The results of this study were also significant in regards to complaints, NOM, DRP and DDI, and to clinical parameters such as systolic blood pressure and body weight.In addition, to some lab exams such as levels of glucose, lipids, thyroid hormones and vitamin D. Silva et al. (2013) showed that the pharmaceutical intervention in dyslipidemia patients taking simvastatin promoted a decrease in the levels of plasma lipoproteins (TC, LDL and TG), a better control of blood pressure and a discrete loss of body weight, corroborating the findings of this study (Silva et al., 2013).Plaster et al. (2012) showed reduced blood pressure, blood glucose and plasmatic lipoproteins in intervention group after 6 months of followup (Plaster et al., 2012). Lyra Júnior et al. (2008) observed that pharmaceutical care intervention optimized the medication use; reduced symptoms caused by drug therapy and improved the elderly patients' health conditions (Lyra Júnior et al., 2008).Castro et al. (2006) showed a trend for better blood pressure control in patients with uncontrolled hypertension after 6 months of follow-up at a pharmaceutical care programs (Castro et al., 2006).Nonetheless, it is possible to note that effectiveness of interventions is more significant in the first six months of follow-up and that a longer period between pharmaceutical appointments diminishes work success.Thus, patients tend to return to baseline conditions.Therefore, there is evidence that it is important to implement pharmacotherapy follow-up in health services as a way to promote better health conditions to the community.
Many factors interfered with executing this study and, consequently, with its effectiveness: availability of space in the workplace, time for appointments (an average of 40 to 60 min per appointment), lack of a fluxogram to attend patients, the staff's lack of experience and knowledge about the work, and doctors' and patients' distrust and suspiciousness.Therefore, to be able to conclude this study, some adjustments in all these issues were necessary.They influenced the sample size and the time needed for development.Some of these obstacles were also identified in the study of Mansoor et al. (2014) and the authors came to the conclusion that time was the main issue (Mansoor et al., 2014).However, one of the sample size's most limiting factors was the number of pharmacists involved in the study (n = 1).Therefore, there was a proportion of one professional to 45 patients (1:45) even though in the initial phases there was a higher proportion.A study by Fikri-Benbrahim et al. (2013) in Spain, for example, involved a proportion of 1 pharmacist to each 10 patients (1:10).A study by Silva et al. (2013) had a proportion of 1:14 (Silva et al., 2013).Most of the studies found involved more than a pharmacist (Castro et al., 2006;Chua et al., 2012;Simpson et al., 2011;Stewart et al., 2014).Ignorance and the lack of custom team to interact with the pharmacist may have influence on the abandonment of this study.The abandonment rate in this study is similar to the others.Balisa-Rocha et al. ( 2014) study showed 58.83% abandonment after 22 months of follow-up.In the study conducted by Correr et al. (2011) only 59.6% concluded the study after 12 months of pharmacotherapy follow-up (Correr et al., 2011).It is believed though, most Brazilian still only demand health services when all are ill and they have no habit in the prevention of health problems.
In the present study, the academic training of the pharmacist, the number of pharmacists to follow-up, the sample size, the group´s age, the time required for each appointment and interventions, the absence of a clinical pharmacy service at the study site, the lack of custom team to interact with the pharmacist, the lack of of patient´s usual to be consulted by a pharmacist, limited resources (medical offices, access to some medical specialties and reagents for lab tests not covered by the covenant) or no resources such as electronic medical records represent the main limitations of this study.Although limited to a geographical region, the results of this study reinforce the importance of the pharmacist being integrated to the health staff, allowing for complete assistance with fewer costs and promoting better health quality to the medication user (Jódar-Sánchez et al., 2014).

Conclusions
The results of this study corroborate literature data that allows us to conclude that identifying DRP and defining strategies to solve them improves the quality of pharmacotherapy and promotes an increase in adherence by the elderly patient.Pharmaceutical interventions allow for a better control of clinical parameters such as blood pressure, body weight, glucose, total cholesterol, LDL, triglycerides, vitamin D and thyroid hormones as the pharmacist works with orienting, monitoring, and encouraging the correct use of medications.Moreover, it works together with other professionals as to encourage healthy life habits.Therefore, pharmacotherapy follow-up promotes a better controle of chronic diseases such as systemic hypertension, diabetes mellitus, dyslipidemia, hypothyroidism and osteoporosis and, consequently, promotes better quality of life to these patients.
The offer of pharmacotherapy follow-up in health services facilities is still modest compared to figures and complications deriving from DRP in dfferent groups.Experience gained from this study allowed for identifying how important this model of pharmaceutical practice is for the community and other health professionals, especially the medical class.In this context, the pharmacist plays a supporting role, in order to guarantee the prescription and the rational use of medications aiming to achieve a higher level of treatment effectiveness and safety for the patient.Many are the challenges for implementing PF in the health services, including training of pharmacy professionals, support and trust from managers, patients and medical class, and financial investments to structure the service and hire more professionals.
Even with those challenges, PF is fully viable to health services because its benefits go beyond the individual plan.By allowing the population to have more access to primary health care, and reducing chronic diseases complications, this model of professional practice promotes fewer hospital admissions, reflecting lower health costs thus, evidencing its collective benefits (Borges et al., 2011;Chisholm-Burns et al., 2010;Jódar-Sánchez et al., 2014;Pittman et al., 2010;Scotti et al., 2013).
a in minimum wage (MW) in Brazil during the data collection period (in reais = R$724.00or in US dollars = US$329) b .

Table 1 .
Number of elderly and abandonment in each phase of the study.

Table 3 .
Main drugs used by the outpatient elderly in pharmacotherapy follow-up, Brasilia-DF (Brazil).

Table 4 .
Main negative outcomes associated with medication in outpatient's elderly in pharmacotherapy follow-up, Brasilia-DF (Brazil).
0.70; CI 95%: 0.50 to 0.82; p < 0.0001 and r = 0.55; CI 95%: 0.30 to 0.72; p < 0.0001, respectively).In other words, the higher is the number of reported complaints by patients, the higher is the percentage of identified ADR and NOM.An analysis of medical prescriptions allowed for identifying 211 drug-drug interactions (DDI) with an average of 4.7 DDI per patient.By using the risk classification, 82.9% (n = 175) of DDI were risk C and 16.6% (n = 35) were risk D. In relation to seriousness, 89.0% (n = 188) of DDI were considered moderate while 10.4% (n = 22) were more serious.The number of drugs taken evidenced a positive and significant correlation with
*Pharmaceutical interventions that require medical approval in Brazil.

Table 6 .
Comparison of pharmacological treatment adherence in baseline and after 12 months of pharmacotherapy follow-up to outpatient elderly, Brasilia-DF (Brazil).

Table 7 .
Evolution of the number of complaints and pharmacotherapeutic variables during 12 months of pharmacotherapy follow-up of outpatient elderly, Brasilia-DF (Brazil).
a Adverse drug reactions; b Negative outcomes associated with medication; c Drug-drug interactions; d p-value obtained by setting mixed effects regression models with linear or quadratic trend;
a Diastolic Blood Pressure; b Systolic Blood Pressure; c Body Mass Index; d p-value obtained by setting mixed effects regression models with linear or quadratic trend;