Drug interactions deserve deeper attention

Treatment guidelines currently in use are generally based on mono-therapies. The debate is open as for the need to consider also drug interactions, a frequent occurrence in polytherapy regimes
Alessandra Marengoni, researcher at the Department of Clinical and Experimental Sciences of the Brescia University (Italy) and member of the Geriatric Working Group (GWG) of the Italian Medicine Agency AIFA

«Until today there has been the lack of an holistic vision of the pharmacological agent, a vision considering not only the disease-specific target of the drug. Each substance may disturb the entire system, especially in the elderly. How to deal with such kind of clinical trials is something very complicate», tells Alessandra Marengoni, researcher at the Department of Clinical and Experimental Sciences of the Brescia University (Italy) and member of the Geriatric Working Group (GWG) of the Italian Medicine Agency AIFA. The study1 on the appropriateness of prescriptions in the elderly published by the Agency in 2013 highlights the central role that should be played by a better evaluation of drug-drug interactions. The study involved the entire over-65 years old Italian population (more than 12 milions people) and was based on the data collected by the National database OsMed for drugs reimbursed by the National Health Service (Table 1). It identified a wide set of appropriateness indicators (table 2) and showed that more than 50 percent of the considered population (7,5 mln) made use of more than 5 drugs per day and 11% (1.4 mln) more than 10.

Alessandra Marengoni, researcher at the Department of Clinical and Experimental Sciences of the Brescia University (Italy) and member of the Geriatric Working Group (GWG) of the Italian Medicine Agency AIFA
Alessandra Marengoni, researcher at the Department of Clinical and Experimental Sciences of the Brescia University (Italy) and member of the Geriatric Working Group (GWG) of the Italian Medicine Agency AIFA


Table 1. Prevalence (%) of quality indicators in the old-age italian population


quality indicator

All ≥ 65 years

(12.301.537 people)

65-74 years

(6.154.421 people)

75-84 years

(4.474.887 people)

≥ 85 years

(1.672.229 people)


–          5/9 medicines

–          ≥ 10 medicines


49,0 %

11,3 %


43,6 %

8,6 %


55,0 %

14,1 %


52,6 %

13,8 %

Low compliance vs anti-depressive drugs 63,9 % 62,6 63,0 69,6
Low compliance vs anti-hypertensive drugs 46,4 43,2 47,2 56,1
Low compliance vs ipoglycemic drugs 63,0 63,0 64,7 70,1
Low compliance vs anti-osteoporosis drugs 52,4 48,7 53,4 64,0
Use of anti-Parkinson e anti-psychotic drugs 0,2 0,2 0,2 0,3
Under usage of statins in diabetic patients

–          % of old population under ipoglycemic treatment









Concomitant use of drugs causing an higher risk of bleeding

–          warfarin + traditional FANS/COX-2

–          warfarin + aspirin/antiplatelet

–          warfarin + traditional FANS/COX-2 + aspirin/antiplatelet

















Concomitant use of drugs increasing an higher risk of renal failure or hyperkalemia (ACE inhibitors/ARB + aldosterone antagonists + FANS/COX-2 inhibitors) 0,7 0,5 0,9 1,0
Concomitant use of ≥ 2 drugs prolonging QT 0,3 0,2 0,4 0,4
Use of anti-hypertensive with an unfavourable risk-benefit profile (doxazosin, clonidine, methyldopa, short time calcium antagonists)

–          % of old population under anti-hypertensive treatment













Use of high dose digoxin (≥ 0,125 mg/die) 0,4 0,3 0,5 1,3
Use of oral antiglycemic drugs assciated with an high risk of ipoglycemia (chlorpropamide, glibenclamide)

–          % of old population under ipoglycemic treatment













(Fonte: adapted from 1J Gerontol A Biol Sci Med Sci (2014) 69 (4): 430-437)



Table 2. Quality of prescription’s indicators from the AIFA study

Polytherapy indicators Concomitant dispensing of:

·         5-9 medicines

·         ≥ 10 medicines

Compliance to treatment ·         Low compliance to anti-depressive medicines (<40% days over 6 months)

·         Low compliance to anti-hypertensive medicines (<40% days over 1 year)

·         Low compliance to anti-diabetic medicines (<40% days over 1 year)

·         Low compliance to anti-osteoporosis medicines (<40% days over 1 year)

Prescriptive cascade Use of anti-Parkinson e anti-psychotic medicines
Under treatment Under usage of statins in diabetic patients
Pharmacological interaction ·         Concomitant use of drugs that may enhance the risk of bleeding (warfarin in combination with traditional inhibitors FANS/COX-2 or low dosage aspirin and other anti-aggregants

·         Concomitant use of drugs that may enhance the risk of kidney failure and/or hyperkalaemia (ACE inhibitors/ARB and aldosterone antagonists, traditional inhibitors FANS/COX-2

·         Concomitant use of more than 2 drugs acting on QT elongation (≥ 2 drugs that may cause Torsades de Pointes)

Medicines to be avoided ·         Use of anti-hypertensive drugs with unfavourable risk-benefit profile (doxazosin, clonidine or methyldopa in monotherapy or any calcium antagonist drug with short action time

·         use of high dosages of digoxin (> 0,125 mg/die)

·         Use of oral ipoglycemic agents associated to an high risk of ipoglycemia (chlorpropanamide o glibenclamide)

(Fonte: AIFA, Uno studio nazionale dell’Agenzia Italiana del Farmaco sulla qualità della prescrizione farmacologica nella popolazione geriatrica, 2013)

The need for a change in clinical trials

The data collected by the study may be more generally extended to the management of all complex patients, no matter for their age. Such kind of patients presents a highly complex pathological framework and calls for the use of polypharmacy regimens of treatment, these too being highly stratified especially in the case of chronic diseases.

«The fundamental issue is that all guidelines for chronic diseases developed in recent years are disease-specific ones. They seldom consider the possibility of co-morbidity with other chronic diseases requiring concomitant treatment – explains Alessandra Marengoni. – These are guidelines based on clinical trials that did not included old co-morbid patients, the one we are dealing with in the daily clinical practice».

Clinical trials are usually based on a disease-specific design in order to better evaluate the safety and efficacy of the drug candidate. The limit of such a design is its inability to highlight the possibility of efficacy bias and side effects related to the interaction of the drug with other medicines or with the mechanisms of the possibly present co-morbid pathologies. These sort of behaviours are usually detected during post-marketing studies, when the new medicine is used by the entire population.

A recent comment published on the British Medical Journal2 By Marengoni and Onder took into consideration the results of the systematic study3 on three NICE Guidelines run by British researchers and aimed to point out drug-disease and drug-drug interactions (“Three nice Guidelines under scrutiny”). «The concomitant applications of the guidelines leads to several interactions – says Marengoni. – In older patients the target is often no more disease-specific. There is the need to take into consideration many more factors and a patient-specific target: its quality of life often prevails. We need to consider the prognosis, the life expectancy and the risk-benefit ratio related to the prescription of a certain regimen. Furthermore, patients are often not sufficiently involved in therapeutic decisions, especially in the elderly. In the case of cognitive or functional impairment, furthermore, there is need for assistance by the family or caregivers».


Three NICE guidelines under scrutiny

The study published in the BMJ took under consideration three different guidelines for the treatment of type 2 diabetes, heart failure and depression which was published by the National Ististute of Health and Care Excellence (NICE). The paper aims to systemically identify, quantify and classify all possible severe drug-disease and drug-drug interactions emerging from the considered guidelines.

First and second line treatment indications have been crossed with those referred to other 11 pathologies. Drug-disease interactions resulted relatively rare, the most severe ones referring to type 2 diabetes. Particularly relevant is the interaction between the prescribed drug and chronic kidney disease.

Much more numerous are drug-drug interactions, as reported in the following table:


Table. Percentage of co-morbid patients under different conditions

type 2 diabete Depression Heart failure Myocardial infarction Chronic kidney diseas Atrail fibrillation COPD Painful conditions Rheumatoidis arthirtis Dementia IHypertension
Type 2 diabetes 18 7 11 14 7 9 23 1 2 61
Depression 7 2 4 4 2 7 27 1 3 23
Heart failure 20 17 36 23 25 18 23 1 1 57

(Fonte: 3BMJ 2015; 350: h949. doi: 10.1136/bmj.h949)


How trials might change

The debate is open on whether to include in clinical trials also old and co-morbid patients in order to better highlight the possible consequences of polypharmacy regimens. A possible solution might reflect the already in place roadmap for approval of pediatric medicines, which requires the mandatory presentation by the sponsor of a Pediatric Investigational Plan (PIP) to the European Medicine Agency (EMA) in parallel with the end of Phase 1 – beginning of Phase 2 studies for each new drug application (EU Regulation n. 1901/2006). «There is need for a more “pragmatic” approach to clinical trials in order to obtain results which would be relevant for daily life. Something is moving, but we are still in a very early phase: after 20-30 years of evidence-based medicine, this approach is no longer considered to be reliable – tells Marengoni. – In 2012, during an EMA workshop with stakeholders, I presented a possible idea to evaluate the effects of combined treatments on specific pathological clusters. The debate is still open on whether these clusters should be the more frequent ones, the more difficult to treat or the more expensives». EU research programmes pay high attention to the identification of new treatment models for co-morbid patients.

The number of over-70 years old people, the new limit of the geriatric age introduced by the World Health Organisation, is continuously increasing. The goal to include complex patients in clinical trials design requires a deep change of the entire vision of drug development and of its use, from research to prescription. «The critical factor is the ability to efficiently communicate to the different stakeholders involved in the treatment of old multi-morbid patients that a disease-specific treatment regimen is no longer acceptable and it might lead to negative effects for the health and quality of life of patients», Marengoni comments.

A support for the general practice

The first point of contact for old and complex patients are often the general doctors, thus they too should be included in the needed change towards patient’s evaluation and prescription behaviour. A main obstacle are the very tight evaluation times general doctors generally have to meet with their patients, often no more than 15-20 minutes. «The cognitive, functional, social and economic situation of the patient should be always taken into consideration – tells Marengoni. – General doctors would need also innovative informatics tools; this sort of instruments are currently different from one another and it is not possible to connect them. It would be very useful to connect the general doctor’s database with the one of the local pharmacy, for example, so to have a double check. Software programmes able to easily and rapidly rise warnings on the main pharmacological interactions, on duplicate drugs, on inappropriate prescriptions, would help to reach the so called personalised medicine».

Adaptive databases

The multidimensional evaluation of the patient may benefit from the development of a new generation of electronic adaptive databases, which can run interactive researches among different pathologies. In a recent editorial on its web site, AIFA stated that this sort of tool, available also on smartphone and tablet, may represent an important piece for the future of the clinical practice and medical profession.

Some examples are already in place, e.g the INTERCheck database on drugs interactions developed by the Mario Negri Institute. The prestigious Italian pharmacology institution also offer a consultancy service by phone for medical professionals and citizens, the Centro di Informazione sul Farmaco e la Salute (CIFS). «I use in my clinical practice this informatics tool to check therapies. Since 2008 we collaborate with the Mario Negri Institute for the Reposi Project, the Italian Register of polytherapies created by the Italian Society of Internal Medicine. I believe it would be important to reach an agreement at the national level between scientific societies and associations, particularly those of the general practitioners and of the pharmacists. It would represent a step forward in order that they themselves would ask for this uniformity», tells Marengoni.

The Norwegian no-profit project MAGIC another example of adaptive database for the creation and distribution of dynamic and continuously updated guidelines, decision supports and summaries of evidence. MAGICapp is the first web-based collaborative platform for the editing and publication of the guidelines: freely available without the need to install any software, it allows the publication on all type of device. «This is not yet enough – says Marengoni. – The patient remains always at the front hedge: it is very important to consider all factors in order to place results in the right context for each individual».


  1. G. Onder et al., High prevalence of poor quality drug prescribing in older individuals: a nationwide report from the Italian Medicines Agency (AIFA), J. Gerontol. A Biol. Sci. Med. Sci. (2014) 69 (4): 430-437, doi: 10.1093/gerona/glt118, first published online: August 2, 2013

  2. A. Marengoni, G. Onder, Guidelines, polypharmacy, and drug-drug interactions in patients with multimorbidity, BMJ 2015; 350: h 1059, doi: 10.1136/bmj.h1059

  3. S. Dumbeck et al., Drug-disease and drug-drug interactions: systematic examination of recommendations in 12 UK national clinical guidelines, BMJ 2015; 350: h949, doi: 10. 1136/bmj.h949



Please enter your comment!
Please enter your name here