The evolution of Reference Drug Lists and Clinical Practice Guidelines in the public health system of a middle-income country

Abstract

Aims

The aims were to analyze the dynamics of the medicines formulary in a middle-income country and to analyze the concordance of the included medicines with the national Clinical Practices Guidelines (CPG).

Methods

Medicines and their indications of use included in the Mexican Reference Drug List (Mex-RDL) from 1996 to 2013 were analyzed. The top 10 indications with the highest number of medicines in 2013 were analyzed retrospectively until 1996 in order to identify the increase in the number of medicines to treat each one, as well as the progressive specificity of the indication according to the International statistical Classification of Diseases (ICD-10). The concordance between the CPG and medicines approved for the top 10 indications was studied.

Results

The number of medicines included in the Mex-RDL kept constantly growing from 454 drugs in 1996 to 811 in 2013. Up to 26.3% of these medicines were approved to treat only 10 indications (1.5% of all possible indications of use). Many of these new medicines had been approved for more and more specific indications, while the oldest ones had been approved for general indications. Up to 27.6% of the medicines approved for these top 10 indications do not appear in the updated recommendations of the specific CPG for those indications.

Conclusions

During the last 18 years, the new medicines and indications included in the Mex-RDL were redundant and concentrated into few similar clinical conditions. This is a factor that promotes an irrational use of these medicines and, thus, unnecessarily raises the price of health care, undermines the quality of the health system and probably increases the uncertainty of treatments.

WHAT IS ALREADY KNOWN ABOUT THIS SUBJECT

• The pharmaceutical industry develops medicines focusing in prevalent and profitable diseases.

• Drug formularies include an oversupply of medicines to treat similar clinical conditions, some of them with little or non-added therapeutic value over the others, but with relevant cost increase.

• Marketing strategies are useful to modulate the medicines demand independently of their rationality.

WHAT THIS STUDY ADDS

• In two decades, more than a quarter of the offer of new medicines in Mexico has concentrated on only 10 health problems or clinical conditions.

• These 10 conditions represent just 1.5% of all possible indications of use of medicines.

• More than one quarter of the medicines to treat these top 10 health problems are not harmonized with the recommendations of the updated Mexican clinical guidelines.

Introduction

The supply of medicines is irrational. On the one hand, there are few or no options to treat some frequent diseases with high health and socio-economic impact [1,2] and on the other hand, there is an oversupply of medicines concentrated in a few diseases or clinical conditions [3–5]. Additionally, the cost of these medicines does not always correlate with their need as a therapeutic option, an optimal risk–benefit relationship or an appropriate use [6,7].

To rationalize the quality and quantity of the medicines offer, the World Health Organization proposed an Essential Medicines List (EML) [8] that established rational criteria to select drugs and the minimum of medicines to be offered. Unfortunately, these goals have not been attained [8,9]. Ideally, the progressive configuration of the market supply should be consequence of the clinical needs according to the epidemiological profile of a given country, but frequently this growth seems to be related to strategies of the manufacturers aimed at modifying prescriber's preferences and patient's consumption of medicines, as well as the performance of the regulatory agencies [10–12].

Manufacturers use different marketing strategies to maximize the uses of medicines and obtain the best benefits, thus modulating the demand of the products independently of their rationality [13,14]. Different studies on marketing strategies and their impact on physicians' and patients' behaviour regarding the prescription and use of medicines identified the conformation of the market supply as an influent marketing strategy [15,16]. With the aim of addressing a potential irrational offer of medicines and marketing influences, some regulatory agencies have implemented reference drug lists and therapeutic guidelines, but sometimes these approaches are not as effective as planned. A recent study carried out in Mexico showed that the Reference Drug List (RDL) used by the Mexican public health system (the ‘Formulario Nacional’) doubled the number of medicines recommended by the EML, but nearly 50% of the analyzed oversupply were medicines without any therapeutic added value [3]. It is plausible to think that this RDL (as others in similar countries) had been partly shaped following marketing strategies.

In 1977 the first edition of the Mexican RDL (Mex-RDL) was published with the aim of improving the use of medicines [17]. The Mex-RDL is mandatory for the whole public health system of the country (that covers 79% of the 112 million of Mexicans) but, not for the private health system (used by 21% of the Mexicans) [18]. Active ingredients are included in the Mex-RDL through a selection process in which an Interagency Commission (with representatives of all public health institutions) assesses the efficacy, safety and cost-effectiveness of the medicines submitted by the pharmaceutical companies [19]. Additionally, the Commission has the mandate to keep updated the Mex-RDL through the inclusion, modification or withdrawal of medicines, at least three times a year [20].

Studies to describe the approved medicines included in a RDL and their authorized indications are useful to identify potential problems that can partly explain inappropriate prescription of medicines. So, the study of the inclusion dynamics for medicines in a formulary could help to understand these marketing strategies and be the basis to design interventions with the aim of improving the utilization of new medicines, trying to avoid prescriptions prompted by novelty and not by clinical need after an accurate risk–benefit analysis.

The present study was carried out to describe these dynamics in the Mex-RDL configuration, as an example of the inclusion process of medicines and their indications. Additionally, the concordance between the Mex-RDL offer and the recommendations of the national Clinical Practice Guidelines (CPG) was analyzed.

Methods

In order to characterize the medicines included in the Mex-RDL, describe the approved indications of use and identify their concordance with the standard recommendations in Mexico, a quantitative and qualitative analysis was designed.

All medicines and their approved indications included or withdrawn in the different editions of the Mex-RDL between 1996 and 2013 were collected and classified [17,19]. Vaccines, hydroelectrolytic solutions and nutritional formulae were not analyzed because they are not medicines.

Analysis of indications

‘Indication’ was defined as any clinical use for a given medicine included in the formularies. So, one medicine can have different indications, and different medicines can share the same indication of use.

All the indications of use approved for each medicine included in the Mex-RDL 2013 were listed. Additionally, the 10 indications with the highest number of medicines in the Mex-RDL 2013 were selected (these were called ‘top 10 indications’). In order to understand how the top 10 indications reached such relevance and how the inclusion dynamics was for the different medicines approved to treat them, these indications were tracked retrospectively until 1996 (the first year with systematic, complete and reliable data available).

For classification purposes, if the indication of use described in the Mex-RDL appeared as an entry under the ‘External Cause’ codes or their subentries in the International Statistical Classification of Diseases and Related Health Problems 10^th Revision (ICD-10) [21], it was considered as a ‘general indication’. Nevertheless, if the precise indication of use did not appear in the ICD-10, it was considered an indication more specialized than a ‘general’ one, and it was referred to as ‘specific indication’. For example ‘breast cancer in premenopausal women’ and ‘breast cancer in post-menopausal women’ or ‘inadequate response to oral hypoglycaemic drugs’ and ‘patients with weight gain that causes problems’ were specific indications of general indications that would be ‘malignant neoplasm of breast’ and ‘diabetes mellitus’, respectively.

Two exceptions were ‘thromboprophylaxis’ and ‘broad-spectrum antibacterial’ as indications, because they could not be classified properly by using an ICD-10 code. Notwithstanding this, due to the high number of approved medicines, as well as their economic impact for populations and institutions, the involved medicines were analyzed. ‘Thromboprophylaxis’ was considered a ‘general indication’ when it was described without any detailed statement, and a ‘specific indication’ when the drug was used for certain types of patients or surgical procedures (e.g. ‘prevention of ischaemic complications in angioplasty’ or ‘percutaneus transluminal atherectomy’, ‘prevention of acute coronary syndrome in percutaneous coronary intervention scheduled’ and ‘thromboprophylaxis in elective total hip replacement and knee or orthopaedic surgery in adults’. In the case of broad-spectrum antibacterials, all medicines were approved with the same clinical use and without specifications regarding diseases. All were classified as ‘general indication’. The proportion of general and specific indications and the number of approved medicines were analyzed year by year.

As the study did not involve patients or medical records, no Ethics Committee approval was requested.

Concordance analysis

The rational use of medicines is based upon different mechanisms [22]. Closed lists of medicines, hospital formularies and CPGs are among the most important. In order to increase their effect, it is important that both are as harmonized as possible (i.e. that the medicines approved for a specific indication match with the medicines recommended in CPGs). Since January 2010, the Mexican Ministry of Health began the review, modification and edition of CPGs for a large number of indications (642 until January 2014) [23]. To assess the concordance between CPGs and the Mex-RDL 2013 recommendations, both information sources were thoroughly reviewed.

Results

The Mex-RDL 2013 included 811 medicines approved for 681 different indications of use. In contrast, the Mex-RDL 1996 included 454 medicines and 690 different indications. That means a 78.6% increase in the number of medicines (n = 357) and a 1.3% decrease in the number of different indications (n = 9) in the last 18 years. During the same period, only 59 medicines with 96 indications of use were withdrawn; so, an average of 20 new medicines was included and three were withdrawn each year.

Medicines to treat the top 10 indications

In 2013, the top 10 indications were: malignant neoplasm of breast (n = 31 medicines; 14.6%;), essential hypertension (n = 30; 14.1%), broad-spectrum antibacterials (n = 25; 11.7%), rheumatoid arthritis (n = 22; 10.3%), type 2 diabetes mellitus (n = 22; 10.3%), asthma (n = 20; 9.4%), thromboprophylaxis (n = 18; 8.5%), generalized idiopathic epilepsy (n = 16; 7.5%;), malignant neoplasm of bronchus and lung (n = 15; 7.0%) and Hodgkin's lymphoma (n = 14; 6.6%).

These top 10 indications of use represented 1.5% of all the 681 possible indications for the medicines included in the Mex-RDL 2013. Notwithstanding, it is noteworthy that up to 26.3% of the medicines included into the Mex-RDL 2013 (n = 213) were approved to treat one of these top 10 indications.

In 1996, these top 10 indications represented 1.4% of the 690 possible indications. Almost half of the 213 medicines for the top 10 indications appearing in the Mex-RDL 2013 (n = 98; 46.0%) were already available in 1996. These 98 medicines represented the 21.6% of the 454 medicines included in the Mex-RDL 1996. That means that 115 new medicines had been included to treat these top 10 indications of use in the last 18 years, 6.4 medicines yearly and a net 117.3% increase (Figure 1). During the same period only seven medicines with 11 indications of use were withdrawn (three broad-spectrum antibacterials, two antihipertensive medicines and two thromboprophylactics).

Figure 1.

Distribution of the medicines in the MEX-LIST 1996 (n = 98) and the Mex-RDL 2013 approved for the top 10 indications of use (n = 213). () 1996, () 2013

General and specific indications

A more detailed analysis of the 213 drugs included in the Mex-RDL 2013 for these top 10 indications showed that 78.4% of them (n = 167) were approved for ‘general indications’ and 21.6% (n = 46) for ‘specific indications’. In 1996, 89.8% of the 98 medicines (n = 88) were approved for ‘general indications’ and 10.2% (n = 10) for ‘specific indications’. Thus, the number of medicines to treat a ‘general indication’ increased by 89.8% (n = 79) and those to treat a ‘specific indication’ increased by 360.0% (n = 36) in the last 18 years.

The growth dynamics of ‘general’ and ‘specific’ indications differed in each one of the top 10 indications. Figure 2 shows the evolution of ‘general’ and ‘specific indications in 1996 and 2013. In 1996, five of the top 10 indications included medicines with ‘specific indications’ (thromboprophylaxis, malignant neoplasm of breast, asthma, essential hypertension and rheumatoid arthritis), but this proportion grew up to eight out of the top 10 indications in 2013. Only the medicines approved for Hodgkin's lymphoma and broad-spectrum antibacterials remained as ‘general indications’.

Figure 2.

‘General’ and ‘specific’ indications of use approved in the Mex-RDL 1996 and the Mex-RDL 2013 for the top 10 indications. () Specific indication, () General indication

It is interesting to analyze how new medicines had been added to each of these top 10 indications along the study period (1996–2013). Figure 3 depicts a breakdown of four of these indications of use and shows how medicines with very specific indications were progressively added to the medicines with ‘general indication’. So, in the case of drugs for malignant neoplasm of breast, most of them were approved as general indications, but medicines to treat ‘metastasis of breast cancer’, ‘breast cancer in premenopausal women’, ‘advanced breast cancer in post-menopausal women’, ‘breast cancer in menopausal women’, ‘breast cancer in women with Her2Neu oncogene’, ‘breast cancer locally recurrent or metastasic’, ‘advanced cancer or metastasis in women with ErbB2 overexpression breast cancer and prior treatment, capacitabine is needed’, and ‘breast cancer in post-menopausal women with locally advanced or metastatic ER+ receptor and progression to prior endocrine therapy’ progressively appeared (Figure 3A). The same can be observed in anti-rheumatoid arthritis medicines (Figure 3B), antidiabetics (Figure 3C) and thromboprophylaxis (Figure 3D).

Figure 3.

Indications of use of the medicines approved to treat malignant neoplasm of breast (A), rheumatoid arthritis (B), antidiabetic medicines (C) and thromboprophylaxis (D) included in different Mexican Reference Lists since 1996 to 2013. A () Post-menopausal women with locally advanced or metastatic ER positive receptor and progression to prior endocrine therapy, () Advanced cancer o metastasis if there is ErB2+ (HER2+) overexpression and prior treatment, trastuzumab included, capecitabine is needed, () locally recurrent or metastatic, () menopausal women, () woman with oncogene Her2Neu, () metastasis, () premenopausal women, () advanced cancer in post-menopausal women, () general indication (malignant neoplasm of breast). B () refractory to DMARDS (Disease-Modifying Antirheumatic Drugs) and one or more biological agents in moderate to severe disease. Should be administered in combination with methotrexate, () resistant to other treatments, () inflammatory joint diseases in general such as rheumatoid arthritis, () general indication (rheumatoid arthritis). C () non-response to metformin alone or patient with linaglipin + metformin controlled but fixed doses is prefered, () inadequate response to two oral hypoglycaemics in patients with body mass index >35 kg m² or before insulin therapy, () non-appropriate control with oral hypoglycaemics (metformin or sulfonylureas) or insulin (alone or with metformin or any sulfonylureas), () inadequate response to oral hypoglycaemics, () non-response to metformin and sulfonylureas, () patients with weight gain ‘that causes problems', when thiazolidinediones are contraindicated or if the patient had a poor response or intolerance to them in the past or in patients who are currently stable with metformin + vildagliptin but fixed doses, () non-response to metformin alone, with or without obesity, () general indication (diabetes mellitus type 2). D () prevention of stroke in patients with non-valvular atrial fibrillation, () prevention of acute coronary syndrome in patients scheduled for percutaneous coronary interventions, () prevention of ischemic complication in angioplasty or percutaneous transluminal atherectomy, () thromboprophylaxis in post-reperfusion coronary patients with thrombolytic agents, () prophylaxis of clotting in the extracorporeal blood circulation, () thromboprophylaxis in elective hip and knee replacement, () general indication (thromboprophylaxis)

Concordance with CPG

The 642 CPGs were carefully reviewed in order to find the concordance between the medicines recommended to treat the top 10 indications in the Mex-RDL 2013. As there is no CPG for Hodgkin's lymphoma (n = 14) in Mexico, the analysis was performed with the 199 medicines approved to treat the remaining nine indications. It is interesting to highlight that 27.6% of the medicines (n = 55) were not included in the equivalent recently updated CPG. Table 1 shows some examples of these discordances.

Table 1.

Medicines included in the Mexican Reference Drug List (Mex-RDL) but not included in the Clinical Practice Guidelines (CPG) for each of the top 10 approved indications of use^† (see Methods)

	In Mex-List	In Mex-List but not in CPG
Indication of use	n	n (%)	Medicines
Malignant neoplasm of breast	31	10 (32.3%)	Idarubicin, ifosfamide, fulvestrant, goserelin, lapatinib, melphalan, mitomycin, testosterone, thiotepa, vinblastine
Essential hypertension	30	11 (36.7%)	Azilsartan, clonidine, isosorbide, telmisartan, valsartan, candesartan + HCT*, irbesartan + HCT*, losartan + HCT*, telmisartan + HCT*, amlodipine + valsartan + HCT*, irbesartan + amlodipine
Broad-spectrum antibacterials	25	1 (4.0%)	Roxithromycin
Rheumatoid arthritis	22	5 (22.7%)	Acetylsalicylic acid, aurothiomalate, dexamethasone, etofenamate, ketoprofen
Diabetes mellitus	22	6 (27.3%)	Exenatide, linagliptin + metformin, liraglutide, lixisenatide, tolbutamide, vildagliptin + metformin
Asthma	20	5 (25.0%)	Epinastine, isoprenaline, terbutaline, zafirlukast (+)
Thromboprophylaxis	18	4 (22.2%)	Abciximab, apixaban, cilostazol, prasugrel
Generalized idiopathic epilepsy	16	3 (18.8%)	Lacosamide, pregabalin, vigabatrin
Malignant neoplasm of bronchus and lung	15	10 (66.6%)	Cyclophosphamide, crizotinib, doxorubicin, erlotinib, gefitinib, ifosfamide, mechlorethamine, mitomycin, pemetrexed, vincristine
Total	199	55 (27.6%)	–

^*HCT, Hydrochlorothiazide; (+) Zafirlukast was approved for two different indications. None of these uses appears into the CPG.

^†In Mexico, there is no CPG for Hodgkins' lymphoma.

Discussion

The present study found that the number of medicines included in the reference list of drugs to be used in the Mexican public health system kept constantly growing at a mean yearly rate of 20 new medicines, from 454 drugs in 1996 to 811 in the 2013 edition of the Mex-RDL. The most striking results are (i) one-quarter of these medicines were approved to treat only 10 indications (1.5% of all possible indications of use), (ii) more than one-quarter of the medicines appearing in the Mex-RDL for these top 10 indications do not correspond with the recommendations included in the recently updated national CPG approved for these indications and (iii) a growing number of these new medicines had been approved for more and more specific indications. All these factors could fuel the inappropriate use of medicines, especially when they are not accompanied by independent continuous education and national campaigns to potentiate rational drug prescription, as is the case of Mexico.

A previous study of the Mex-RDL content in 2010 showed a two-fold irrationality in that some essential medicines recommended by the World Health Organization to treat prevalent diseases were missing, and medicines without any rational added value were in oversupply [3]. Medicines oversupply seems to be a worldwide phenomenon driven by the manufacturers who push their products into the medicines market probably facilitated by permissive or loose regulations. This panorama allows the marketing of new medicines not driven by their need (according to what is already marketed, efficient and in use for a specific indication), but by commercial purposes [4–7]. Thus, the present study was designed to analyze the evolution of the inclusion of new medicines in a middle-income country during the last 18 years. Additionally, due to the need that a formulary is accompanied by the corresponding CPGs, an analysis of the concordance between the medicines approved to treat the top 10 indications of use and those recommended in the CPGs was performed.

Inclusion of new medicines and update of the reference list

In 2013, the Mex-RDL contained 811 medicines to cover the needs of more than 112 million inhabitants. The mean number of new medicines included by year during the study period was 20, a figure similar to that found in other countries [4,24,25]. Unfortunately, this tendency is not accompanied by the withdrawal of obsolete medicines, thus leading to an endless and steady growth of the list [17,19]. Curiously, despite this increase in the number of medicines, the number of qualitatively different indications of use, showed a slight decrease, thus resulting in more new medicines concentrated in already existing indications. This is particularly worrying, because physicians who began their clinical practice 18 years ago, should have learnt (either by themselves or through continuous education programmes) pharmacokinetics, pharmacodynamics, efficacy and toxicity information about 357 new medicines not explained during their undergraduate courses in order to be able to carry out the selection process necessary for a good prescription. Unfortunately, the health system rarely offers updating and continuous education programmes for prescribers in order to provide independent and unbiased information that help to go beyond the purely marketing messages [13,26].

Towards the rationality of the market width: the need of harmonized tools

Perhaps the equilibrium between the medicines marketed in Mexico and those included in a country formulary intended for the public health system such as the Mex-RDL should be desirable, and the way to reach this equilibrium would be a reasonable concordance between the well-based decisions of the Therapeutic Committees (TC) to include or exclude a particular medicine and the contents of independent and periodically reviewed CPGs. To study the concordance between the medicines approved to treat the top 10 indications of use in the Mex-RDL and the recommendations of the guidelines for these 10 indications of use was one of the objectives of the present research. Obviously, the fact that a medicine is included in a CPG does not imply that this medicine is used rationally, but being there, at least improves this probability. Common sense suggests that if there is the possibility to prescribe a medicine in the public health system (because it is included in the Mex-RDL), but this medicine is not recommended in the CPG of that country, the chances for an inappropriate use of that product or an inappropriate treatment of that disease, greatly increase [22].

The study described herein showed that 27.6% of the 199 medicines approved for the top 9 indications are not recommended to treat these indications in the corresponding CPG. A detailed review of these disparities and the reasons that led to the present situation would help to improve the situation in Mexico and to improve the use of medicines. In the previous 3 years the National Center for the Excellence Technology in Health (NCETH), a branch commission of the Ministry of Health, has been updated or written 642 CPGs. This process followed a high standard methodology to select and discuss the different therapeutic options, and included some of the most recognized clinical experts of the different public health institutions [23]. In fact, the lack of harmonization between the offer of the Mex-RDL (that depends on the national TC) and the CPG (that depends on the NCETH) translates a problem common to several health systems: the absence of coordination and collaboration between institutions [27,28]. The present results add more evidence to propose a joint assessment in order to fill the gaps or reorder the offer of medicines. Such an approach could help to improve the use of medicines.

Evolution of the indications of use

One of the most striking observations of the present study is the fact that one quarter of the Mex-RDL medicines were approved to be prescribed in only 1.5% of all the possible indications of use (rop 10 indications). Although this situation is well-known, the present study was an opportunity to quantify it. Such ‘concentration’ of medicines competing for a few health problems can be explained by the incidence and characteristics of the involved diseases. Hypertension, diabetes mellitus, breast cancer, rheumatoid arthritis or the need for anticoagulation are, at the same time, prevalent diseases and conditions that are considered good marketing opportunities [12,29,30].

Additionally, a 360% increase in the specification of use of the approved medicines was also observed during the study period, but specifically in the last years. To request the inclusion of a new medicine for a new and highly specific indication probably has more chances to result in a positive answer than to ask for the inclusion of a new medicine for an indication that already has a vast number of alternatives to be selected for.

In Mexico, the Interagency Commission approves the inclusion of new medicines according to some well-established criteria that do not include a limit in the number of medicines by indication [19,20]. The Mexican drug policy promotes the theoretical idea that an increase in the number of therapeutic choices will stimulate price competition (i.e. the appearance of cheaper medicines) [10]. In fact, these conditions allow the approval of more and more medicines with similar risk–benefit, but in oligopoly markets such as the Mexican one, price elasticity is almost non-existent [31].

In other countries, the approval rate of the TCs is between 47.0 to 88.4% of the applications [32–35]. The observed oversupply [3] as well as the concentration of a lot of options for very few health problems probably do not help to promote a rational use of medicines. Although it is difficult to know how much is ‘excessive’ in the case of medicines and health problems, it has been described that an oversupply is associated with an increase of the inappropriate treatments and the appearance of adverse drug reactions [36]. In this context, there are a couple of parallel discussions without any answer at the moment: (i) how many different options should there exist to cover each indication of use? and (ii) what is an ‘indication of use'? The Mexican medicines' regulations lack explicit criteria about these questions, thus the deeper the medical knowledge on a disease, the more specific indications are described and the more medicines are approved. At this moment, both the Interagency Commission and the NCETH are very active and prone to improve the access to efficient, safe, cost-effective and quality medicines, so it should be desirable to begin the discussion on these topics in order to bridge the existing gaps [23,35]. To harmonize these two instruments in order to align the official recommendations and the supply of medicines in the public health system would certainly improve the use of medicines.

The results of the present study should be interpreted taking into account several limitations. On the one hand, it is an observational study carried out in Mexico, a very specific Latin American medicines market, and some considerations are necessary before extrapolating the conclusions of the research to other health systems or countries. Notwithstanding, the increase in the medicines approved during the last two decades and the idea to focus more and more the indications of use in order to try to create a ‘therapeutic gap’, as well as the strategies to push a new medicine into the market are common practices everywhere, especially in the present globalization era [12,37]. On the other hand, it is logical that the more evolved the medical, physiological and pharmacological knowledge, the more can be detailed the alleged mechanism of action of the medicines. Notwithstanding, it should be kept in mind the number of ‘new’ mechanisms of action claimed for medicines that have been sold as real novelties fated to mark a milestone, but ended with important label alerts or even quick drug withdrawals due to risk–benefit imbalances that were not well evaluated. A few recent examples (e.g, dabigatran among anticoagulant medicines [38], bevacizumab in breast cancer [39,40], rofecoxib in arthritis [41] or rosiglitazone in diabetes mellitus [42,43]) should be remembered and used for educational purposes.

Studies of drug formularies or reference drug lists usually are focused on the medicine's risk–benefit profile, costs and inappropriate or off-label uses, but not frequently in their indications of use. In the present study we tried to turn the table through approaching the medicine uses in more realistic conditions, and this could be considered one strength of the research. Studies such as the present one could be carried out in other countries with similar epidemiological profile, health system coverage, income health budget and permissive or loose legislation in order to identify common patterns of medicines approval and the marketing strategies applied for. Considering the present pharmaceutical pipelines in therapeutic areas such as schizophrenia, cancer, depression, dementia or glaucoma, specific indications in these fields are to be expected [12,37].

This knowledge could help to strengthen medicines authorities and improve their regulatory decisions according to well-established criteria such as efficacy, safety, convenience and cost, thus improving the quality of the medicines offer in a country [44]. It should be noted that some recent changes in the inclusion process have been done, but there is room for additional improvement (e.g. limitation of the number of products per indication, inclusion of new products only when advantages over an already effective medicine have been shown, or taking into account the ‘need’ factor for the approval of a new medicine, as well as the periodical review and withdrawal of useless drugs). Additionally, the description of inconsistencies between the reference list and the recommendations of the CPGs are a necessary step to reduce the prescription of medicines not based on the best available evidences.

The appropriate use of medicines is a sign of quality in healthcare, a way to improve patient healing and to limit unnecessary adverse drug effects. This appropriate use depends on many factors and actors along the life cycle of a given medicine. A critical and independent review of the medicines included in a reference list, as well as their approved indications of use, together with their inclusion in clinical practice guidelines are actions that could help to improve the quality of the medicines market, and the quality of the health care.

Competing Interests

There are no competing interests to declare. All authors have completed the Unified Competing Interest form at http://www.icmje.org/coi_disclosure.pdf and declare: no support from any organization for the submitted work, no financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years and no other relationships or activities that could appear to have influenced the submitted work.