European Journal of Clinical Pharmacology
Eur. J. Clin. Pharmacol. 17, 25-31 (1980)
© by Springer-Verlag 1980
Hospital Admissions Due to Adverse Drug Reactions: A Comparative Study from Jerusalem and Berlin M. Levy 1, H. Kewitz 2, W. Altwein 2, J. Hillebrand 2, and M. Eliakim 1. 1 The Clinical Pharmacology Unit, Department of Medicine A Hadassah University Hospital, Jerusalem, Israel, and z Institute of Clinical Pharmacology, Klinikum Steglitz, Free University of Berlin, Germany
Summary. A comparative study of adverse drug reactions (ADR) leading to hospital admission showed that 103 (4.1%) out of 2499 medical admissions in Jerusalem and 167 (5.7%) out of 2933 admissions in Berlin were due to such reactions. Sex distribution in the two patient - populations was almost equal but the Jerusalem patients were younger. The most frequent ADRs were digitalis intoxication (in Berlin) and reactions to antibiotics (in Jerusalem). Other important differences were noted in the relative frequencies of ADRs associated anticoagulants, hypoglycemic agents and oral contraceptives. They were probably related to differences in drug usage in the two countries. The most common major side effects were arrhythmias, allergic reactions, bleeding, congestive heart failure, bronchospasm and hypoglycemia. The following risk factors were identified in both cities: old age, female sex, impaired renal function, previous history of A D R and polypragmasia.
Key words: adverse drug reactions, digitalis intoxication, antibiotic reactions; drug utilization, hospital admissions, clinical adverse effects
Hospital admissions due to conditions classified as adverse drug reactions (ADR) are an index of the magnitude, severity and type of iatrogenic disease in outpatient populations [1-4]. Geographical variation in the rate and nature of adverse reactions may be due to genetic, environmental and demographic factors, as well as to different patterns of drug utilization and medical practice. * Established Investigator, Chief Scientist's Bureau, Ministry of Health, Israel.
The present paper compares the results of studies of hospital admissions due to A D R in general medical wards in Jerusalem [5] and in Berlin. An analysis of the drugs involved, the clinical character of the reactions, identification of risk factors and possible explanations for some of the differences between the two study-populations are presented.
Subjects and Methods Both the Department of Medicine A of the Hadassah University Hospital in Jerusalem and the Institute of Clinical Pharmacology, Klinikum Steglitz of the Free University of Berlin have participated in the Boston Collaborative Drug Surveillance Program (BCDSP). A detailed description of the methodology of the BCDSP has previously been published [6-8]. The present study is based on 2499 patients in Jerusalem admitted during 1969-1976, and 2933 patients in Berlin admitted during 1974-1977. The Jerusalem patients were selected at random and they represented about 20% of all subjects admitted to the department during that period. In Berlin, all patients admitted to four out of five medical wards were included. The admission policy in the Berlin hospital differed from that in Jerusalem in two major points: 1) exclusion of patients admitted to the coronary care unit in Berlin, and 2) patients with a cutaneous reaction in Berlin were usually hospitalized in the Department of Dermatology. For every case admitted, the attending physician was asked by a specially trained nurse-monitor, whether the admission could possibly have been due to an ADR. All such cases were later re-evaluated by clinical pharmacologists. A D R was defined according to the criteria of the American Registry of Pathology of the Armed Forces Institute of Pathology [9]. Only 0031-6970/80/0017/0025/$01.40
26
M. Levy et al.: Comparison of Adverse Drug Reactions
Table 1. Distribution of patients admitted for A D R and for other reasons, by age and sex (%) Jerusalem Admitted for adverse drug reactions (n = 103)
Admitted for other reasons (n = 2396)
Age groups (years)
Male
Female
Total
Male
~<20 21-30 31-40 41-50 51-60 61-70 71-80 >81
3.9 2.9 1.9 4.9 5.8 12.6 2.9 0
2.9 11.7 9.7 3.9 13.6 16.5 5.8 1.0
6.8 14.6 11.6 8.8 19.4 29.1 8.7 1.0
5.3 7.5 5.1 6.8 7.9 11.1 6.6 2.1
3.8 7.2 5,3 6.1 8,3 9.5 6,0 1,4
9.1 14.7 10.4 12.9 16.2 20.6 12.6 3.5
3.3 4.1 4.6 2.8 4.9 5.7 2.9 1.1
total
35.0
65.0
100.0
52.4
47.6
100.0
4.1
Female
Rate of admission for A D R (%) Total
Berlin Admitted for adverse drug reactions (n = 103)
Admitted for other reasons (n = 2396)
Age groups (years)
Male
Female
Total
Male
Female
<20 21-30 31-40 41-50 51-60 61-70 71-80 >81
0.6 1.2 3.0 3.6 5.4 7.8 12.6 3.0
1.2 5.4 2.4 2.4 7.8 12.5 21.0 10.1
1.8 6.6 5.4 6.0 13.2 20.3 33.6 13.1
1.3 4.0 5.4 5.6 6.4 12.0 11.7 4.7
1.2 3.7 7.0 6.8 6.4 12.2 9.6 2.1
2.5 7.7 12.4 12.4 12.8 24.2 21.3 6.8
4.1 4.9 2.6 2.8 5.9 4.8 9.5 10.4
total
37.1
62.9
100.0
51.1
48.9
100.0
5.7
definite and probable reactions, representing a major factor leading to hospitalization, were considered. Patients who were admitted because of attempted suicide were excluded, but reactions after blood transfusion (given on a day-care basis), or following biologicals (e. g., rabies vaccine) were included. Data collected in all cases included vital statistics, diagnoses, blood urea nitrogen (BUN), liver function tests on admission and drugs taken during the month preceding hospitalization. An inquiry was made into adverse drug reactions in the past. Computer analysis of the data from Jerusalem was performed at the BCDSP Center in Boston and of those from Berlin by the Zentraleinrichtung ffir Datenverarbeitung Dialogsystem Siid. The chisquared test was used for statistical analysis. Results
103 (4.1%) out of 2499 patients hospitalized in Jerusalem between 1969 and 1976 and 167 (5.7%) out of 2933 patients hospitalized in Berlin during the years 1974-1977 were admitted because of an ADR.
Rate of admission for A D R (%) Total
Sex and Age (Table 1) Females predominated among patients admitted for A D R (65%; p < 0.001) in Jerusalem, and 62.9% (p < 0.001) in Berlin). Of the total patient population in Jerusalem 53% were over 50 years of age, whereas in Berlin 65% were in this age group. Although there was no linear correlation between age and the rate of admission for ADR, the rate in Berlin for patients under 50 y was 3.3% and in those over 5 0 y it was 6.9%. The relative numbers of women patients in Berlin admitted for A D R compared to patients admitted for all other reasons was twice as great in the age group 71-80 years and fivefold greater in patients aged 80 years and above.
Diagnoses (Table 2) The most common discharge diagnoses in Jerusalem were ischemic heart disease, congestive heart failure, diabetes mellitus, hypertension and respiratory infections, and in Berlin, ischemic heart disease, congestive heart failure, gastrointestinal disease and diabetes mellitus. In spite of some differences, such as
M. Levy et al.: Comparison of Adverse Drug Reactions
27
the higher incidence of rheumatic heart disease in Jerusalem, the diagnoses in the two patient-populations were quite comparable (Table 2).
Drugs Suspected of Causing ADR (Tables 3 and 4) About 60 and 80 preparations given in Jerusalem and in Berlin, respectively, were identified as suspected causes of ADR. Single drugs caused the admission of 84 (82%) of the patients in Jerusalem, and of 129 (79%) of the patients in Berlin. All other admissions were related to drug combinations. In Jerusalem, eight drugs were responsible for 41 Table 2. Common Discharge Diagnoses (% of all admitted patients) a
Congestive heart failure Diabetes mellitus Acute respiratory infectionb Gastrointestinal disease Rheumatic heart disease Malignant tumors Neurological disease
Jerusalem
Berlin
11.3 10.9 9.1 5.8 5.0 4.2 3.7
13.9 6.3 6.2 8.8 2.2 5.8 2.8
a Ischemic heart disease and hypertension have not been included because of differences in classification. b including pneumonia.
(39.7%) of all admissions for ADR. Drugs most frequently responsible for hospitalization were cardiac, antibiotic, analgesic and antineoplastic agents. In Berlin 73 (44%) of all admissions for A D R were associated with eight preparations, amongst which cardio-active drugs, analgesics, hypoglycemic agents, anticoagulants and oral contraceptives predominated. Digitalis glycosides were the most likely cause
Table 3. Drugs involved in admissions for adverse drug reactions (% of admissions due to ADR)
Drug classes
Jerusalem
Digitalis glycosides fi-blockers Antiarrhythmics (Quinidine) Antibiotics Analgesics Antineoplastics Corticosteroids Diuretics Antihypertensives Anticoagulants Antidiabetics Oral contraceptives Hypnotics Others and combinations Total
7.7 4.9 4.9 16.5 9.7 9.7 7.7 5.8 4.9 2.9 1.0 1.0 1.0 22.3 100
Berlin 34.1 2.4 0 1.2 16.8 2.4 4.2 3.0 1.2 6.6 10.8 4.7 1.2 11.4 100
Table 4. Commonest drugs leading to hospitalization in comparison with two other studies (% of all reactions) Present study
Previous studies
Drug
Ampicillin Digoxin Furosemide Quinidine Acetylsalic. acid t-blockers Melphalan Digitoxin Dipyrone Prednisone Blood Guanethidine Warfarin Spironolactone Insulin Indometacin Hydrochlorothiazide Compound analgesics Oral antidiabetics
Jerusalem
Berlin
Miller d [3]
Caranasos [4]
7.7 6.8 4.9 4.9 4.9 4.8 3.9 3.9 3.9 3.9 2.9 2.0 1.0 1.0 1.0 0 0 0 0
0 27.5 a 0 0 1.2 2.4 0 3.0 0.6 2.4 b 1.2 0 6.6 c 2.4 6.6 1.8 0 12.0 4.2
0.4 17.3 0.4 1.1 11.0 0 0.6 0.4 0.4 5.9 0 2.0 5.4 0 0 0 1.1 0 0
unknown 14.2 1.8 0 14.1 0 unknown 1.0 0 2.8 0 0.4 6.8 0 0 0 4.0 0 0
including derivatives b including prednisolone c in Berlin - phenprocoumon was used as anticoagulant d 14% of the data came from Jerusalem patients
M. Levy et al.: Comparison of Adverse Drug Reactions
28 Table 5. Major manifestations of ADR leading to hospitalization
(% of all reactions) Jerusalem No % Blood dyscrasia with or without bleeding Rash and urticaria Cardiac arrhythmia GI bleeding Congestive heart failure Fever Orthostatic hypotension Respiratory disease Anaphylactic shock Electrolyte imbalance Renal failure Hypercorticism Angina pectoris Thrombophlebitis Other Hypoglycemia Nausea, vomiting Total
21 17 10 6 5 4 4 3 3 3 3 3 2 2 17 0 0 103
20.4 I6.5 9.7 5.9 4.8 3.9 3.9 2.9 2.9 2.9 2.9 2.9 1,9 1.9 16.5 0 0 100
Berlin No 21 3 37 22 0 2 0 5 0 4 1 6 0 5 24 17 20 167
% 12.6 1.8 22.1 13.2 0 1.2 0 3.0 0 2.4 0.6 3.6 0 3.0 14.3 10.2 12.0 100
in about one third of the admissions in Berlin, and in about 8% of those in Jerusalem. Marked differences between the two populations were noted for antibiotics, hypoglycemic drugs, anticoagulants, oral contraceptives and antiarrhythmics. In terms of individual drugs, ampicillin, furosemide and quinidine were common suspected causes of A D R in Jerusalem, but no such cases were noted in Berlin. On the other hand, digoxin, insulin, phenprocoumon and oral hypoglycemics were relatively frequently associated with A D R in Berlin.
Type of Reaction (Table 5) Allergic rash, bone marrow depression, arrhythmias, bleeding and congestive heart failure accounted for the admission of 59 (57.7%) of all patients admitted for A D R in Jerusalem. In Berlin, arrhythmias were followed in descending order of frequency by gastrointestinal bleeding, blood dyscrasias, vomiting and hypoglycemia. These reactions constituted 70% of all admissions for ADR. The relatively small number of
Table 6. Patients admitted for life-threatening ADR A: Jerusalem Number of patients
Implicated drug (s)
Adverse drug reactions
1-2 3 4 5-6 7 8 9 l0 11
Digoxin Digoxin and Furosemide Digitoxin and Furosemide PractoloI, Propranolol Methimazol Gold Dipyrone Blood Fluorescein
Ventricular tachycardia, pulmonary edema Hypokatemia, V. E. S., Stupor Trifascicular block Pulmonary edema Agranulocytosis Exfoliative dermatitis, sepsis Anaphylactic shock Anaphylactic shock Anaphylactic shock
Digoxin Digitoxin Insulin Gtibenclamid Phenprocoumon r-blockers (Pindolol, Propranolol, Metoprolol) Spironolactone Acetylicsalicylic acid Chloroquine Methimazol Spalt® Spalt* + Dipyrone Penicillamine, Naproxen Vincristine, Cyclophosphamide Vincristine, Adriblastine Azathioprine Asthma-Kranir~ Clonidine
Severe arrhythmias and conduction disturbances Severe arrhythmias and conduction disturbances Severe hypoglycemia Severe Hypoglycemia Melena, hematemesis (5 cases), CVA (1 ease) Severe bronchospasm (3 cases) pulmonary edema (1 case) Hyperkalemia (7.1-8.3 meq/1) Thrombocytopenia (5,000, 12,000) Agranulocytosis Agranulocytosis Thrombocytopenia (4,000) Thrombocytopenia (8,000) Thrombocytopenia (34,000) Leukopenia (1,400), bronchopneumonia Leukopenia (900), Thrombocytopenia (59,000) Thrombocytopenia (2,600) and bleeding Respiratory arrest After withdrawal - hypertensive crisis with severe encephatopathy
B: Berlin 1-13 14 15-25 26-30 31-36 37-40 41-42 43-44 45 46 47 48 49 50 51 52 53 54
M. Levyet al.: Comparisonof Adverse Drug Reactions admissions in the Berlin series due to drug rash is probably due to the fact that such cases were hospitalized in a dermatological ward. Similarly, in Jerusalem patients with complicated arrhythmias were hospitalized in the intensive coronary care unit, and those with massive gastrointestinal bleeding in the surgical department.
Outcome There were no deaths among patients admitted because of A D R in Berlin. In Jerusalem, however, five patients (4.9% of the admissions for A D R and 0.2% of all admissions) died as a result of suspected ADR. Three patients died from aplastic anemia following the administration of chloramphenicol for a trivial infection, one patient died from intracranial hemorrhage due to thrombocytopenia following quinidine 1.2 gram daily for 21 days, and one patient developed bone marrow aplasia and fatal sepsis following a one week course of nitrogen mustard. Eleven patients (10.7% of all admissions for A D R ) in the Jerusalem series had reactions sufficiently severe to be considered life-threatening, i. e. considered likely to be fatal without immediate treatment (Table6). Cardiac drugs were most often involved (six cases). Life-threatening A D R represented 0.5% of all admissions to the department. In Berlin, 54 patients (32% of the admissions for A D R and 1.8% of all admissions) had a lifethreatening A D R . Sixteen of them were suspected to be due to antidiabetic agents (insulin 11, glibenclamide 5), 14 to digitalis glycosides and 6 to phenprocoumon.
Risk Factors In addition to female sex the following factors were evaluated:
(a) History of Adverse Drug Reaction. A positive history was obtained in 43 (41.7%) patients admitted for A D R in Jerusalem, and in 26.8% of patients admitted for other reasons. In Berlin 54 of the cases (32%) reported an A D R in the past, whereas only 2.3% of all other monitored patients remembered having had a drug reaction in the past. Sixteen patients (15.5%) in Jerusalem reported that the same type of A D R that led to admission had already occurred in the past, while in Berlin 28 (16.8%) patients had such a history. The causes included digitalis intoxication, insulin-hypoglycemia, furosemide-induced gout, nitrofurantoin-induced allergy, melena associated with corticosteroids and leukopenia due to cytotoxic agents.
29 TaMe 7. Drugs taken one month prior to hospitalization(% of all admitted patients) Jerusalem Berlin I: Analgesics (simple) Dipyrone Compound codein Other compound analgesics Aspirin Acetaminophen II: Cardiacs Digoxin Quinidine Propranolol Digitoxin Other beta-blockers III: Antibiotics Ampicillin Tetracycline Chloramphenicol IV: Diureticsand hypotensives Furosemide Chlorothiazide Methyldopa Spironolactone Guanethidine V: Antidiabetics Insulin Oral hypoglycemics VI: Anticoagulants Heparin Oral anticoagulants VII: Others Oral contraceptives Cytotoxics Prednisone and prednisolone Diazepam Meprobamate
21.5 13.6 12.8 11.9 5.9
3.0 1.5 23.4 2.6 0.1
13.8 4.5 4.3 2.3 1.3
38.8a 0.3 3.5 2.3 1.9
12.7 9.5 2.5
1.5 1.1 0.1
12.8 6.2 5.3 3.2 0.5
5.9 5.3 2.3 7.8 0.2
2.2 5.7
3.2 9.5
0.3 2.2
0.2 1.7
1.5 2.0 5.2 14.8 6.0
3.8 1.7 5.4 11.0 0.1
a includingacetyl and methyl derivatives
(b) Number of Drugs Taken During the Month Preceding Hospitalization. Patients who were admitted for A D R had taken more drugs during the month preceding hospitalization than had the comparison groups. The average number of drugs taken by the patients admitted for A D R was 4.9 in Jerusalem, 4.25 in Berlin and 3.1 in both comparison groups. The drugs most commonly taken by all patients prior to hospitalization are shown in Table 7. They should be compared with Table 4 which lists the drugs responsible for hospitalization. It is clear from this comparison that there is some correlation between the types of drugs used prior to admission and those responsible for admission, although this is not true for all drugs. (c) Impaired Renal Function. In the Jerusalem series, a BUN value of 21 mg% or more was found in 46.6% of patients admitted for A D R and in 30.3% of the patients admitted for other conditions. In Ber-
30
lin, 41% of patients admitted for A D R had a plasma creatinine level exceeding 1.0 mg%, while the corresponding figure in patients with other conditions was 32%. Impaired renal function could therefore be identified in both cities as a risk factor.
(d) Impaired Liver Function. There was no correlation between elevated serum bilirubin or SGOT level and A D R in either study population. Discussion
The rates of admission for conditions suspected to be ADRs (4.1% in Jerusalem and 5.7% in Berlin) are within the range reported by others - 1.7-5.6% [1-4]. Such conditions are one of the most common causes of medical admissions, and so represent a significant medical and economical problem [10]. Although it is quite impossible to prevent A D R in general, some of the drug reactions could have been avoided since they had occurred in the past and could have been predicted had a more careful drug history been taken. The total number of drugs involved in admissions for A D R (60 and 80 in Jerusalem and Berlin, respectively) represent only a small fraction of the drugs marketed in both countries. Thus, the major ADRs in the outpatient population are caused by a limited number of widely used, potent, allergenic or toxic drugs, which demand careful individualization of dosage because of their pharmacokinetic properties. The recognition of risk factors, such as impaired renal function and polypragmasia, is also of the utmost importance in minimising ADR. A discussion of the role of predisposing factors for A D R has previously been presented [5]. Differences in the rate and character of admissions due to A D R reported in previous studies may be explained by variations in methodology, monitoring techniques and definitions of A D R [11], as well as to genuine discrepancies between the study populations. In the present study, differences in methodology were minimized, and so the differences found are likely to be inherent to the two populations and/or to varying patterns of drug utilization. A possible explanation for the somewhat higher rate for A D R in Berlin is the relatively older age of that population. It is well known that morbidity is greater, polypragmasia more common, and drug responsiveness and disposition are altered in the aged [12-15]. Specifically, the significantly higher rate of digoxin intoxication in Berlin can be explained by several of these factors. Digoxin is more frequently used by older patients, who are also more susceptible
M. Levy et al.: Comparison of Adverse Drug Reactions
to intoxication than younger subjects because of decreased clearance of the drug [16]. Although 11.3 and 13.9% of the patients in Jerusalem and Berlin, respectively, suffered from congestive heart failure (Table2), 39% of the total number of patients admitted in Berlin were reported to have used digoxin during the month preceding hospitalization (Table7), whereas in Jerusalem only 13.8% had done so. The widespread use of digoxin in Berlin may account for the large number of patients admitted with reactions associated with this drug. In addition, the poor bioavailability of the Israeli brand of digoxin used at the time of the study [17] may be partly responsible for the relatively low rate of digoxin intoxication in Jerusalem. It is of interest that the frequency of A D R to digitoxin, a drug for which problems of bioavailability or impairment of renal function are not important, and the rate of usage of which was similar in the two populations, was almost the same in Jerusalem and Berlin. Quinidine was rarely reported to have been used during the month preceding hospitalization in Berlin, whereas in Jerusalem 4.5% of the patients reported its administration. Consequently, 4.9% of all admissions due to A D R in Jerusalem were associated with quinidine, whereas in Berlin this drug was not listed amongst those which were considered of having caused admissions. On the other hand, the relatively higher rate of ADR associated with contraceptives in Berlin may have been due to the lower consumption of those drugs in Jerusalem, as reflected in Table 8. Analgesics were thought to have been responsible for 9.7% of A D R in Jerusalem and for 16.8% in Berlin. Dipyrone, the common analgesic in Israel and in Germany (in compound preparations), is virtually unused in the USA, where aspirin is the leading drug in the group. This is reflected in the relatively low rate of A D R related to aspirin in this study as compared to studies from the USA (Table 4) [3, 4]. Two of the reactions associated with dipyrone noted in the two centers were considered to have been lifethreatening (thrombocytopenia and anaphylactic shock), but only one patient was admitted with leukopenia. This finding is in contrast with the allegedly high rate of dipyrone agranulocytosis suspected in the USA [18]. Although analgesic drugs are extremely widely used, the urgently needed comparative studies of the relative risks involved in their administration are not yet available. Further differences between the two cities were found with respect to A D R associated with hypoglycemic drugs - 11% in Berlin and only 1% in Jerusalem. Diabetes was a more common diagnosis in Jerusalem, but a history of antidiabetic drug use during the month preceding hospitalization was more
M. Levy et al.: Comparison of Adverse Drug Reactions
common in Berlin (12.7%) than in Jerusalem (7.9%). This difference is not likely, therefore, to account for the tenfold higher rate of servere hypoglycemia leading to hospitalization. Utilization rates of anticoagulants were similar in both centers, yet the rate of admission for A D R to anticoagulants was much higher in Berlin. The explanation for the higher rate of A D R associated with these drugs in Berlin is probably related to different admission policies, usage patterns or control of effect of these drugs. The different rates of A D R associated with antibiotics are mainly due to the high rate of admission of patients with reactions associated with ampicillin in Jerusalem; ampicillin was much more commonly used by the Jerusalem population (Table 8). In addition, Berlin patients with urticaria and other skin manifestations were usually admitted to the dermatological ward, thus accounting for the "lower" rate in that city. Chloramphenicol was considered to have been responsible for three deaths due to bone marrow depression in Jerusalem. The Ministry of Health in Israel has since restricted the use of chloramphenicl only to hospitals. This complication was not observed in Berlin, possibly because the study there started in 1974, when physicians were already well aware of the risk involved. It seems that in many cases the difference in the relative distribution of drugs suspected of leading to hospitalization for A D R reflect important differences in the pattern of drug utilization in the two countries. The recognition of differences in therapeutic practice is important, as it may lead to a comparative study also of the efficacy of therapy, for example, prevention of vascular complications in diabetes mellitus and the protective effects of anticoagulants [19].
Acknowledgement. The
Drug Monitoring Program in Berlin has been supported by a grant from the Bundesminister f/Jr Jugend, Familie und Gesundheit der Bundesrepublik Deutschland, Bonn-Bad Godesberg. References 1. Smith, J.W., Scidl, L.G., Cluff, L.E.: Studies on the epidemiology of ADR. V. Clinical factors influencing suceptibility. Ann. Int. Med. 65, 629-60 (1966)
31 2. Hurwitz, N.: Admissions to hospital due to drugs. Br. Med. J. 1969/1, 539-540 3. Miller, R. R.: Hospital admissions due to adverse drug reactions. Arch. Int. Med. 134, 219-223 (1974) 4. Caranasos, G.J., Stewart, R.B., Cluff, L.E.: Drug-induced illness leading to hospitalization. J. Am. Med. Assoc. 228, 713-717 (1974) 5. Levy, M., Lipshitz, M., Eliakim, M.: Hospital admissions due to adverse drug reactions. Am. J. Med. Sci. 277, 49-56 (1979) 6. Borda, I., Slone, D., Jick, H.: Assessment of adverse reactions within a drug surveillance program. J. Am. Med. Assoc. 205, 645-647 (1968) 7. Slone, D., Jick, H., Borda, I., et al.: Drug surveillance utilizing nurse monitors. Lancet 1966/1I, 901-903 8. Jick, H., Miettinen, O.A., Shapiro, S. et al.: Comprehensive drug surveillance. J. Am. Med. Assoc. 213, 1455-1460 (1970) 9. Irey, N. S.: Diagnostic problems and methods in drug-induced diseases: Part I (1966), PartII (1967), PartIII (1968). Washington, DC: American Registry of Pathology, Armed Forces Institute of Pathology, 1966-1968 10. Mach, E. P., Venulet, J.: The economic of adverse reactions to drugs. WHO Chronicle 29, 79-84 (1975) 11. Karch, F., Lasagna, L.: Adverse drug reaction. J. Am. Med. Assoc. 324, 1236-1241 (1975) 12. Hurwitz, N.: Predisposing factors in AR to drugs. Br. Med. J. 1969/I, 536-539 13. Levy, M., Nir, I., Birnbaum, D., et al.: Adverse reactions to drugs in hospitalized medical patients. Isr. J. Med. Sci. 9, 617-626 (1973) 14. Gorrod, J. W.: Absorption, metabolism and excretion of drugs in geriatric subjects. Gerontol. Clin. 16, 3 0 4 2 (1974) 15. Crooks, J., O'Malley, K., Stevenson, I.H.: Pharmacokinetics in the elderly. Clin. Pharmacokinet. 1, 280-296 (1976) 16. Ewy, G.A., Kapadia, G. G., Yao, L., et al.: Digoxin metabolism in the elderly. Circulation 39, 449-453 (1969) 17. Levy, M., Adler, R., Superstein, E., et al.: Biovailability of digoxin tablets in Israel. Harefuah 89, 153-154 (1975) 18. Huguley, C.M.: Agranulocytosis induced by dipyrone, a hazardous antipyretic and analgesic. J. Am. Med. Assoc. 189, 935-941 (1964) 19. Lunde, P. K. M., Levy, M.: Drug utilization - geographical differences and clinical implications. In: Advances in pharmacology and therapeutics, Vol. 6. P. Duchene-Marrulaz (ed.), pp. 79-81. Oxford: Pergamon Press 1978
Received: July 11, 1979 accepted in revised form: October 23, 1979
Prof. Dr. H. Kewitz Institut fiir Klinische Pharmakologie der Freien Universitfit Berlin Klinikum Steglitz Hindenburgdamm 30 D-1000 Berlin 45