Adverse drug reactions in elderly patients: how to predict them?

246

Passarelli MCG, Jacob Filho W

Original ARTICLE

Adverse drug reactions in elderly patients: how to predict them? Reações adversas a medicamentos em idosos: como prevê-las? Maria Cristina Guerra Passarelli1, Wilson Jacob Filho2

ABSTRACT Objective: Adverse drug reactions (ADR) present great epidemiological importance among elderly people because of their high prevalence and potential for complications. This study had the objectives of determining the risk factors for ADR in a hospitalized elderly population and, from these, to create an instrument that would enable their prediction. Methods: The study population consisted of 186 elderly people (≥ 60 years) hospitalized in the internal medicine ward of a teaching hospital. Using an active search method, the patients were assessed daily for investigating and diagnosing ADR. The associated risk factors were obtained by means of a multiple logistic regression model and, from these, the backward variable selection method was used to create an ADR Prediction Instrument. Results: Among the population evaluated, 115 patients (61.8%) presented at least one ADR (91% were type A). The risk factors considered significant for ADR were the number of diagnoses (OR = 1.41; 95% CI: 1.06-1.86), number of drugs (OR = 1.10; 95% CI: 1.03-1.17) and drug use that was inappropriate for elderly people (OR = 2.32; 95% CI: 1.17-4.58). From these data, a cutoff value for the number of drugs utilized could be calculated, such that if the patient used a number of drugs greater than this, ADR risk greater than 0.5 would be presented. Conclusions: Application of the ADR Prediction Instrument will make it possible for patients at greater risk to be monitored while using drugs, thus improving care quality. Keywords: Elderly people; Hospitalization; Pharmaceutical preparations/ administration and dosage; Risk factors; Damage prediction

RESUMO Objetivo: Reações adversas a medicamentos (RAM) apresentam grande importância epidemiológica em idosos por sua elevada prevalência e potencial para complicações. Este estudo teve como objetivo determinar os fatores de risco para RAM em uma população idosa hospitalizada e, a partir destes, criar um instrumento que permita a sua previsão. Métodos: A população do estudo foi composta de 186 idosos (≥ 60 anos) internados na enfermaria de clínica médica de um hospital-escola. Com um método de busca ativa, os pacientes foram avaliados diariamente para pesquisa e diagnóstico de RAM. Os

fatores de risco associados foram obtidos por meio de um modelo de regressão logística múltipla e, a partir deles, adotou-se o método de seleção de variáveis backward para a criação de um instrumento de previsão de RAM. Resultados: Entre a população avaliada, 115 pacientes (61,8%) apresentaram, no mínimo, uma RAM, sendo 91% do tipo A. Os fatores de risco considerados significativos para RAM foram o número de diagnósticos (OR = 1,41; IC 95%: 1,06-1,86), o número de medicamentos (OR = 1,10; IC 95%: 1,03-1,17) e o uso de medicamento inapropriado para idosos (OR = 2,32; IC 95%: 1,174,58). Com base nesses dados, é possível calcular o valor de corte do número de medicamentos empregados, de tal forma que, se o paciente usar um número de medicamentos superior a esse valor, apresentará risco de RAM superior a 0,5. Conclusões: A aplicação do instrumento de previsão de RAM permitirá que pacientes sujeitos a maior risco sejam monitorizados enquanto usarem medicamentos, aprimorando a qualidade do cuidado. Descritores: Idoso; Hospitalização; Preparações famacêuticas/ administração e dosagem; Fatores de risco; Previsão de danos

INTRODUCTION Although drug therapy is paramount to control most diseases, it is widely known that there is no medication that is totally safe, since all drugs, in a greater or lesser extent, may trigger an Adverse Drug Reaction (ADR), which is defined as a “noxious and unintended response to the use of a medication, associated with the use of a dose generally employed in human beings for prophylaxis, diagnosis and treatment of diseases and/or to change physiologic functions, excluding the cases of therapeutic failure”(1). The epidemiological importance of ADR is justified for its high prevalence rate – they cause from 3% to 6% of hospital admissions at any age, and up to 24% in the elderly population; they rank fifth among all causes of death and, moreover, they represent from 5 to 10% of hospital costs(2).

Study carried out at Municipal Hospital Centre of Santo André, State of São Paulo, Brazil (teaching hospital linked to ABC School of Medicine). PhD, Lecturer of Internal Medicine and Propedeutics at the Faculdade de Medicina do ABC – FMABC, São Paulo (SP), Brazil.

1

Full professor of Geriatrics, Medical School, Universidade de São Paulo – USP, São Paulo (SP), Brazil.

2

Corresponding author: Maria Cristina Guerra Passarelli – Rua Gaivota 916 – ap. 91 – Moema – CEP 04522-032 – São Paulo (SP), Brasil – Tel.: 11 5055-4460 – e-mail: [email protected] Received on Dec 18, 2006 – Accepted on Jun 13, 2007

einstein. 2007; 5(3):246-251

Adverse drug reactions in elderly patients: how to predict them?

Elderly citizens are especially vulnerable to ADR because of factors related to this age group: pharmacokinetic and pharmacodynamic features, presence of multiple diseases (co-morbidities), use of a large number of drugs (polymedication) and the type of prescribed medication (proper or inappropriate). We must highlight that certain drugs or classes of drugs should not be prescribed for elderly individuals, because they present high risks of causing severe drug reactions, there is insufficient evidence of benefits and there are therapeutic options that are more, or as effective as, and pose lower risks. Such medication, called inappropriate for the elderly, are defined by consensus among specialists, following the Beers criteria(3), very much employed in the last 15 years in clinical practice, in research and continuing medical education. These criteria are available at the site of the Health Surveillance Center of the State of São Paulo: www.cvs.saude.sp.gov.br. Adverse reactions are ranked according to their severity as severe, which are life threatening, the ones that cause or prolong hospital stay, or those that bring about permanent or significant sequelae; and non-severe(1). As to their mechanism, they are classified in: A- much more common and prevalent in the elderly, expected, dosedependent and related to the very pharmacological effects of the drugs; and B- more rare and associated with a high degree of morbidity and mortality, not dose-dependent or dependent on the known pharmacologic properties of the medication, frequently immunomediated or genetically-based(4). In conclusion, the possibility of preventing an adverse reaction would be a highly desirable measure, avoiding its occurrence or allowing for early identification and treatment, thus reducing the impacts of this important cause of morbidity and mortality.

OBJECTIVES To determine the risk factors for ADR in a sample of elderly hospitalized patients and, from then on, to create prevention tools. METHODS This study was carried out at the Internal medicine Ward of the Centro Hospitalar Municipal de Santo André (Faculdade de Medicina do ABC – University Hospital), from September of 2002 to May of 2004, after being approved by the Research Ethics Committee of the institution. All elderly patients (≥ 60 years) admitted during this period were considered eligible to enter the research protocol, and we excluded the individuals who had conditions that prevented data collection – such as

247

intubated patients under mechanical ventilation and/or in coma – as well as those who had already been to other hospital departments for at least seven days before being admitted to the Internal Medicine ward, because in such cases it would not be possible to have an accurate diagnosis of their eventual prior adverse reactions. For each patient enrolled in this investigation, we filled out a protocol in which we collected data such as age, sex, schooling, history of alcohol intake, history of prior adverse reactions, diagnoses, list of all medications used and their doses, list of complementary exams, description of any events or complications, results of the Mini Mental State Examination and the Scale of Daily Activities and dates of hospital admittance, hospital discharge and/or death. By means of using an active search method(5), the patients were assessed daily for study and diagnosis of adverse reactions to medication. We tried to carefully investigate each symptom, sign or laboratory alteration, aiming to identify the highest possible number of potential ADR, and for that we used specific reference material(6-7). Each possible ADR was assessed according to the degree of likelihood, following the Naranjo algorithm(8). The reactions identified were classified according to severity as severe or non-severe and, according to their mechanism: A or B. All medications used were characterized as being adequate or not, based on Beers criteria – 2003 version(3). It is also worth mentioning that the medications were considered inappropriate regardless of base diagnosis or associated diseases.

Statistical analysis We carried out a descriptive analysis by means of building, for the quantitative variables, tables with mean values, standard deviations, maximum and minimum values; and for the qualitative variables, tables with frequency and percentage distributions. Mean values were compared according to variance analysis techniques or Student’s t test, observing whether the variances were equal or not, and data normality. Proportions were compared by means of the chi-squared test. Risk factors for ADR were determined by means of adjusting a stepwise multiple logistics regression model to select the variables. Following that, a new adjustment was carried out, using the backward variable selection method, in order to create a tool used to predict ADR likelihood. RESULTS Of the 269 elderly patients admitted during the study period, 74 (17.9%) were excluded since it was not possible to collect their data and 9 (2.2%) refused to participate. einstein. 2007; 5(3):246-251

248


Therefore, the population in the study encompassed 186 elderly – whose main characteristics are described in Table 1. One hundred and fifteen of them (61.8%) had at least one adverse reaction to one drug during their hospitalization, totaling 199 ADR, with a mean of 1.7 reactions per patient (minimum of 1, and maximum of 8). Of all the adverse reactions observed, 73 were considered severe (36.7%) and 126 non-severe

(63.3%), as shown in Table 2. We must emphasize that some ADR showed up in both lists, which is explained by the definition of severity and, therefore, by the very context of each reaction. The type-A reactions, predictable on the basis of the drug pharmacological profile, prevailed (181, or 91%) over type-B (18 or 9%). Table 3 shows the most prescribed medication and their association with ADR.

Table 1. Main characteristics of the elderly population studied (patients with ADR* vs. with no ADR*) Characteristic Age, %

ADR

60-69 70-79 ≥ 80

Female, % Alcohol consumption, % Falls, % Past history of ADR, % Use of inappropriate drug, % Dependence to carry out Activities of Daily Life, % Alteration in the Mini Mental Examination, % Schooling (years) 0 1-3 ≥4 Renal failure, % Hepatic failure, % Number of diagnoses (mean ± SD; median) Number of drugs (mean ± SD; median) Hospital stay (days) (mean ± SD; median)

p value

Yes

No

38 38 24 52 10 36 67 76 38 77

38 37 25 62 18 32 70 54 31 82

23 40 37 9 5

35 37 28 7 4

3.2 ± 1.2;3

2.6 ± 1.2;2

0.003

13.6 ± 7.5; 13

9.9 ± 4.5; 9

< 0.001

20.0 ± 14.2; 15

10.4 ± 5.6; 10

< 0.001

0.953

0.191 0.126 0.650 0.622 0.002 0.314 0.405 0.201

0.687 0.759

* Adverse drug reactions

Table 2. Severe and non-severe adverse reactions Severe ADR Acute renal failure Hyperkalemia Orthostatic hypotension Digitalis intoxication Acute atrial fibrillation Digestive hemorrhage Hypoglycemia Gastritis Dehydration Diarrhea Paroxysmal supraventricular tachycardia Others Total Non-severe ADR Hypokalemia Skin rash Hypotension Somnolence Epigastralgia (upper gastric pain) Dehydration Diarrhea Vomiting Hypoglycemia Hyperglycemia Nausea Cough First degree atrioventricular block Headache Others Total

einstein. 2007; 5(3):246-251

Frequency 13 7 6 5 4 4 4 3 2 2 2 21 73 Frequency 21 10 9 9 7 6 6 6 5 4 4 4 3 3 29 126

(%) 17.8 9.6 8.2 6.8 5.5 5.5 5.5 4.1 2.7 2.7 2.7 28.7 100.0 (%) 16.7 7.9 7.1 7.1 5.6 4.8 4.8 4.8 4.0 3.2 3.2 3.2 2.4 2.4 23.0 100.0

The most significant risk factors in relation to adverse reactions, according to the multiple logistic regression model, were the number of diagnosis (OR = 1.41; CI 95%: 1.06-1.86), the number of drugs (OR = 1.10; CI 95%: 1.03-1.17) and the use of inappropriate medication by the elderly (OR = 2.32; CI 95%: 1.17-4.58)(9). The model is properly adjusted (Hosmer-Lemeshow test: p = 0.348). Based on the adjusted logistics regression model used, a rule was created – called ADR prediction tool for the elderly – and used to predict the risk of an elderly patient having an adverse reaction, based on the number of diagnoses, the number of drugs used, and the use of inappropriate medication. By this rule it is possible to determine the cutoff value of the number of drugs employed, in such a way that, if the individual used a number of medications above this value, he/she will have an ADR risk above 0.5. The cutoff value depends on the number of diagnoses and the fact of the patient using or not inappropriate medication. Table 4 shows the cutoff values obtained for different numbers of diagnoses, in the cases in which the elderly takes or does not take inappropriate medication. These values were obtained from the logistics regression model, setting the ADR possibility as 0.5.


Table 3. Frequency of use of most prescribed medications Medications (inappropriate in bold font) Captopril Acetylsalicylic acid Furosemide Heparin Ranitidine Ceftriaxone Ipratropium bromide Fenoterol Dipyrone Diazepam Rapid acting insulin Omeprazole Spironolactone Digoxin* Potassium chloride Metoclopramide Aminophylline Deslanoside Ciprofloxacin Propranolol Hydrochlorothiazide Isosorbide mononitrate Amiodarone Glibenclamide Hydrocortisone Propatyl nitrate Clindamycin Chlorthalidone Erythromycin NPH insulin Nifedipine Enoxaparin Methyldopa Cimetidine Diclofenac Promethazine Chlorpropamide Fluoxetine Amitriptyline Dexchlorpheniramine Clonazepam Meperidine Mineral oil Tenoxicam Nitrofurantoin Oxybutynin

Number and percentage of patients who used medications and number and percentage of patients who presented an ADR Total of * % Patients with % patients ADR 138 74.1 25 18.1 108 58.6 12 11.1 95 51.0 24 25.3 93 50.0 3 3.2 79 42.4 0 0.0 77 41.4 8 10.4 73 39.2 0 0.0 70 37.6 10 14.3 62 33.3 0 0.0 48 25.8 5 10.4 46 24.7 3 6.5 46 24.7 1 2.2 45 24.1 3 6.7 43 23.1 8* 18.6 41 22.0 0 0.0 37 19.8 0 0.0 36 19.3 8 22.2 36 19.3 1 2.8 35 18.8 5 14.3 35 18.8 5 14.3 34 18.2 3 8.8 33 17.7 3 9.1 32 17.2 3 9.4 32 17.2 4 12.5 31 16.6 12 38.7 31 16.6 1 3.2 27 14.5 2 7.4 26 13.9 7 26.9 24 12.9 4 16.7 24 12.9 4 16.7 24 12.9 2 8.3 23 12.3 0 0.0 20 10.7 4 20.0 15 8.06 1 6.7 12 6.4 2 16.6 8 4.3 2 25.0 7 3.7 0 0.0 7 3.7 0 0.0 5 2.6 1 20.0 5 2.6 0 0.0 4 2.1 0 0.0 4 2.1 0 0.0 2 1.0 0 0.0 2 1.0 1 50.0 1 0.5 1 100.0 1 0.5 0 0.00

∗According to Beers criteria, digoxin is considered inappropriate for elderly when used in dosage higher than 0.125mg/day – it occurred in 7 out of 8 patients with digitalis intoxication.

Table 4. Tool to predict ADR* in the elderly Use of inappropriate drugs Number of diagnosis 1 2 3 4 5

No 18 15 11 7 4

Yes 9 6 2 0 0

*Adverse drug reactions

For example, we noticed that, if the patient does not take inappropriate medication and if the number of diagnosis is equal to 2, the patient will be then assigned to the group of higher risk of developing adverse reactions if the number of medication taken is higher than 15. On the other hand, if he/she uses inappropriate

249

medication and the number of diagnoses is higher than 3, he/she will be included in the group of higher risk for developing adverse reactions, regardless of the number of medication taken. This rule was used for the elderly in our study, and 70.4% of them were correctly classified (sensitivity = 88.7% and specificity = 40.8%).

DISCUSSION Despite being an important public health problem, adverse drug reactions are are frequently underdiagnosed and not properly appreciated. Nonetheless, it has been broadly shown that they represent an important cause of hospitalization, and especially in the elderly population, they prolong hospital stay, increase medical care costs and could even cause death – of which prevalence varies between 2 and 5% of all patients with ADR(2,1013) . Another undesirable consequence of ADR is the possibility of it being interpreted as a real disease, leading to unnecessary investigations and new risks inherent to diagnostic procedures. Many of the adverse reactions observed in this investigation are frequently found in medical practice, nonetheless, it is true that they are not always identified as ADR. Furthermore, according to the literature, the drugs described as the most prevalent in the present investigation – ACE inhibitors, antibiotics, diuretic agents, hypoglycemic agents and non-hormonal antiinflammatory drugs – seem to be the ones most often associated with adverse reactions(13-15). In fact, all treatment modalities that involve the use of medication carry an implicit risk of causing ADR. Type-B reactions, regardless of drug pharmacological properties, are considered unpredictable. Nonetheless, most adverse reactions, in the general population, and most specifically in the elderly, correspond to type-A, related to the drug pharmacological properties and, exactly for this reason, are considered predictable and, thus, preventable(4). In this study, the findings of highest prevalence of type-A reactions, are corroborated by the literature(16-17). Therefore, due to high morbidity and mortality rates related to ADR and the possibility of being able to predict it for most of the cases, we believe that each and every effort should be geared towards preventing them, or at least identify them early on, thus limiting their undesirable consequences. Many authors have proposed approaches to reduce ADR in the elderly, and almost all of them are related to medical education and better prescription habits. Thus, some actions are proposed: the preferential use of drugs with high therapeutic index (18) , the precise knowledge about pharmacology of the drug einstein. 2007; 5(3):246-251

250


chosen (19), the use of the fewest possible drugs in carefully calculated doses(20) and avoiding to prescribe medication that may cause drug interaction(21), among others. Also, another constant finding in the literature is the importance given to the early identification of patients under a higher risk of developing ADR and the need for their strict monitoring(13,22). Nonetheless, in our careful literature review, we did not find any reference as to some type of tool that, based on the existence of risk factors, could be systematically applied to predict ADR. The three risk factors that in this investigation presented a significant association with adverse reactions seem to have a very coherent relation among each other. In fact, it is reasonable to suppose that a patient with many diagnosis should require many drugs and, therefore, would run a greater risk of receiving an inappropriate drug. The finding of using the number of diagnosis as a predicting factor for adverse reaction development is in agreement with the reports from many authors, and it is a well established risk factor(2,23-24). As to the use of many drugs, this is the major risk factor for ADR – as it is described in most studies similar to the present one(2,22-23,25-27). The presence of inappropriate drugs in the prescription has been considered a frequent cause of adverse reactions in the elderly, and it even doubles the risk(9,28-30). Therefore, based on these three highly relevant and significant risk factors, according to the literature, we intended to create a tool that may be useful to control ADR in the elderly and also to create best practices of medical prescription. However it was applied later on, in such a way as to point out which patients would be prone to a more careful monitoring. Please, notice that there are three key questions to be asked: how many diagnoses the patient at hand has? How many drugs have been prescribed? Based on Beers criteria, is any drug considered inappropriate? Using new examples, imagine that a hypothetic patient has only one diagnosis and does not use inappropriate medication; in such a situation, only in case this patient’s prescription has more than 18 drugs will his ADR risk be greater than 0.5. If, on the other hand, this patients has 5 diagnoses, and he needs to use an inappropriate medication, his risk for developing ADR will be greater than 0.5, regardless of how many drugs were prescribed. All individuals with a similar profile should be carefully monitored during their entire hospital stay, by reassessing the indication for each one of the drugs prescribed, their exact dose calculation, checking medication serum level – when applicable, and a constant surveillance for early detection of each and every ADR that may manifest itself despite all care taken. After diagnosing an ADR, one einstein. 2007; 5(3):246-251

must decision as to which is the best course of action: discontinue the use of the drug involved, reduce its dose or associate another drug, the latter is only used when the first drug is really required.

CONCLUSION By using this tool, we suggest that patients under a greater risk of developing ADR should be closely monitored while using medication which doses and indications require accurate calculations and assessments. It is also worth mentioning the importance of considering ADR in the differential diagnosis of any new symptom, sign or laboratory abnormality presented by the patient. Thus, this would be an interesting and feasible way to enhance the quality of care provided to elderly patients. REFERENCES 1. Edwards IR, Aronson JK. Adverse drug reactions: definitions, diagnosis and management. Lancet. 2000;356(9237):1255-9. 2. Onder G, Pedone C, Landi F, Cesari M, Della Vedova C, Bernabei R, et al. Adverse drug reactions as cause of hospital admissions: results from the Italian Group of Pharmacoepidemiology in the elderly (GIFA). J Am Geriatr Soc. 2002;50(12):1962-68. 3. Fick DM, Cooper JW, Wade WE, Waller JL, Maclean R, Beers MH. Updating the beers criteria for potentially inappropriate medication use in older adults. Arch Intern Med. 2003;163(22):2716-24 . 4. Meyboom RHB, Lindquist M, Egberts ACG. An ABC of drug-related problems. Drug Saf. 2000;22(6):415-23. 5. Laporte JR. Farmacovigilancia en el hospital. In: Laporte JR, Tognoni G, editors. Principios de epidemiología del medicamento. 2a ed. Barcelona: Masson; 1993. p. 219-31. 6. Dukes MNG, Aronson JK, editors. Meyler’s side effects of drugs. 4th ed. New York: Elsevier Science; 2000. 7. Davies DM, Ferner RE, de Glanville H, editors. Davies’s textbook of adverse drug reactions. 5th ed. London: Chapman and Hall Medical; 1998. 8. Naranjo CA, Busto U, Sellers EM, Sandor P, Ruiz I, Roberts EA, et al. A method for estimating the probability of adverse drug reactions. Clin Pharmacol Ther. 1981;30(2):239-45. 9. Passarelli MCG, Jacob Filho W, Figueras A. Adverse drug reactions in an elderly hospitalised population: inappropriate prescription is a leading cause. Drugs Aging. 2005;22(9):767-77. 10. Classen DC, Pestotnik SL, Evans RS, Lloyd JF, Burke JP. Adverse drug events in hospitalized patients: excess lenght of stay, extra costs and attributable mortality. JAMA. 1997;227(4):301-6. 11. Bordet R, Gautier S, Le Louet H, Dupuis B, Caron J. Analysis of the direct cost of adverse drug reactions in hospitalized patients. Eur J Clin Pharmacol. 2001;56(12):935-41. 12. Pirmohamed M, James S, Meakin S, Green C, Scott AK, Walley TJ. Adverse drug reactions as cause of admission to hospital: prospective analysis of 18.820 patients. BMJ. 2004;329(7456):15-9. 13. Atkin PA, Veitch PC, Veitch EM, Ogle SJ. The epidemiology of serious adverse drug reactions among the elderly. Drugs Aging. 1999;14(2):141-52. 14. González-Martín G, Yáñez LC, Valenzuela EA. Reacciones adversas a medicamentos en pacientes geriátricos hospitalizados. Estudio prospectivo. Rev Med Chile. 1997;125(10):1129-36. 15. Hutchinson TA, Flegel KM, Kramer MS, Leduc DG, Kong HH. Frequency, severity


and risk factors for adverse drug reactions in adult out-patients: a prospective study. J Chron Dis. 1986;39(7):533-42. 16. Lagnaoui R, Moore N, Fach J, Bégaud B, Longy-Boursier M. Adverse drug reactions in a department of systemic diseases-oriented internal medicine: prevalence, incidence, direct costs and avoidability. Eur J Clin Pharmacol. 2000;56(2):181-6.

251

23. Carbonin P, Pahor M, Bernabei R, SgaADRi A. Is age an independent risk factor of adverse drug reactions in hospitalized medical patients? J Am Geriatr Soc. 1991;39(11):1093-9. 24. Veehof LJG, Stewart RE, de Jong BM, Haaijer-Ruskamp FM. Adverse drug reactions and polipharmacy in the elderly in general practice. Eur J Clin Pharmacol. 1999;55(7):533-6.

17. Chan M, Nicklason F, Vial JH. Adverse drug events as a cause of hospital admission. Int Med J. 2001;31(4):199-205.

25. Courtman BJ, Stallings SB. Characterization of drug-related problems in elderly patients on admission to a medical ward. Can J Hosp Pharm. 1995;48(3):161-6.

18. Routledge PA, O’Mahony MS, Woodhouse KW. Adverse drug reactions in elderly patients. Br J Clin Pharmacol. 2004;57(2):121-6.

26. Roughead EE, Gilbert AL, Primrose JG, Sansom LN. Drug-related hospital admissions: a review of australian studies published 1988-1996. Med J Aust. 1998;168(8):405-8.

19. Walker J, Wynne H. Review: the frequency and severity of adverse drug reactions in elderly people. Age Ageing. 1994;23(3):255-9. 20. Goettler M, Schneeweiss S, Hasford J. Adverse drug reaction monitoringcost and benefit considerations. Part II: cost and preventability of adverse drug reactions leading to hospital admission. Pharmacoepidemiol Drug Saf. 1997;6(Suppl 3):S79-90. 21. Seymour RM, Routledge PA. Important drug-drug interactions in the elderly. Drugs Aging. 1998;12(6):485-94. 22. van den Bent PM, Egberts AC, Lenderink AW, Verzijl JM, Simons KA, et al. Risk factors for the development of adverse drug events in hospitalized patients. Pharm World Sci. 2000;22(2):62-6.

27. Wu WK, Pantaleo N. Evaluation of outpatient adverse drug reactions leading to hospitalization. Am J Health-Syst Pharm. 2003;60(3):253-9. 28. Onder G, Landi F, Cesari M, Gambassi G, Carbonin P, Bernabei R. Inappropriate medication use among hospitalized older adults in Italy: results from the Italian Group of Pharmacoepidemiology in the Elderly. Eur J Clin Pharmacol. 2003;59(2):157-62. 29. Pitkala KH, Strandberg TE, Tilvis RS. Inappropriate drug prescribing in homedwelling elderly patients. Arch Intern Med. 2002;162(15):1707-12. 30. Lindley CM, Tully MP, Paramsothy V, Tallis RC. Inappropriate medication is a major cause of adverse drug reactions in elderly patients. Age Ageing. 1992;21(4):294-300.

einstein. 2007; 5(3):246-251