Skip to main content
Download PDF

Application of STOPP/START criteria in older patients in primary care using RStudio®

BMC Geriatrics volume 24, Article number: 782 (2024) Cite this article

Abstract

Objectives

To estimate the frequency of potentially inappropriate prescribing (PIP) in outpatients according to STOPP/START criteria, and to identify risk factors. For this purpose, an algorithm was developed and validated in RStudio® based on the information collected in the electronic prescription.

Methods

The data corresponds to dispensations from two pharmacies in Spain made to patients over 18 years, over 4 years. For the analysis, only patients aged ≥ 65 years who are targeted by the STOPP/START criteria are included. The statistical programming language RStudio® was used to develop the algorithm. The STOPP criteria used as models for the implementation of the method were L2 and B12. A logistic regression analysis was performed.

Results

A total of 15,601 treatment plans were obtained from 2312 patients ≥ 65 years (56% women), of whom 46.6% had polypharmacy (≥ 5 drugs) and 9.3% had excessive polypharmacy (≥ 10 drugs). In this group, PIPs were detected in 57% of patients and in 38% of their treatment plans; of these PIPs the most common were those related to the use of benzodiazepines for more than 28 days (D5 criterion) in 25.9% of patients, followed by the use of opioids prescribed without an associated laxative (L2 criterion) in 13.8% and finally, drugs duplication (A3 criterion) in 5.7%. The most numerous duplications related to criterion A3 were benzodiazepines (39%) and non-steroidal anti-inflammatory drugs (37.4%).

Conclusions

The method developed and validated in RStudio® with different STOPP criteria allows us to analyse the pharmacological treatment of many patients using different databases and to identify those at risk of suffering a PIP according to the STOPP criteria.

Our results indicate a high prevalence of PIPs in patients ≥ 65 years, with polypharmacy being the most common risk factor affecting PIP.

Peer Review reports

Background

The world's population is ageing: most countries in the world are experiencing an increase in the number and proportion of older people. As of January 2022, the population aged over 65 years in Spain amounted to 9.54 million people, thus continuing the upward trend witnessed in previous years [1]. Between 2002 and 2022, the elderly population has increased by more than two million. According to recent data, people aged over 65 years represent nearly a fifth of the Spanish population [1, 2] It follows that the future of the healthcare system is marked by the progressive ageing of the population and therefore by the chronicity of pathologies. The consequences go beyond the simple sum of illnesses because there is also a greater consumption of medicines and a higher frequency of polypharmacy [3]. All this makes it necessary to develop initiatives for the management of chronic patients.

Different authors highlight the prevalence of potentially inappropriate prescriptions (PIPs) in older patients and underline the relationship between PIPs and adverse drug reactions, hospitalizations, morbidity, and inefficient use of healthcare [4,5,6]. The STOPP/START criteria [7], among others, have been used as tools to facilitate the identification of PIPs in older patients. The version used in this work was version 2, published in 2015. These standards are regularly revised, and the most recent one was just released [8].

Several studies have been carried out, both in Europe and in the USA, using different methods to assess PIPs and their prevalence, reaching values ranging from 10 to 70% depending on the health care setting, the study method, and the population [4,5,6,7, 9]. Thus, in the field of primary care, the works of Cahir et al. [10], Bradley et al. [11] Díaz-González et al. [12] and Fuertes Abardía et al. [13] stand out. Cahir et al. [10] and Bradley et al. [11] studied the prevalence of PIP in the population aged 70 years and older using the 2008 STOPP/START criteria. Both studies assessed the association between PIP and number of medications, gender and age, finding that about [10]35% of the total had some PIP and of these, about 70% had only one PIP, between 12–22% had two PIPs and finally, 8% were prescribed three or more PIPs [11].Díaz-González et al. [12] analysed PIP in polypathological patients aged ≥ 65 years in a health centre, comparing the STOPP-START criteria (2009 and 2015, versions in Spanish). The 2015 update of the criteria detects more PIPs than its predecessor (2009 version); 85% of people with any STOPP or START criteria, compared to 63% respectively. Finally, Fuertes Abardía et al. [13] analysed PIP in the elderly using the 2015 STOPP/START criteria (version 2). They assessed the association of PIP with several variables (polymedication, sex, age and socio-demographic characteristics). According to the results obtained, PIP is higher in women, institutionalised and home patients, and in those with a higher number of drugs and number of visits.

In the current healthcare context, it is increasingly important to prioritise patient interventions according to the time and resources available [6]. Assuming that prescription errors in patients aged 65 and over are attributable to variables that can be identified with the information available in the electronic prescription, the aim of this work is to rapidly detect infringements of the STOPP/START criteria, as well as to identify the variables associated with the infringement of these criteria. To this end, an algorithm is developed, implemented and validated in RStudio® that allows the analysis of the information available in the electronic prescription.

Methods

Participants

According to official data for the year 2022, 17.1% of the population of the Canary Islands is 65 years of age or older. Of the total population, 49.4% are men and the average age is 43 years [14]. For the detection of STOPP/START criteria, information from outpatient electronic prescription dispensing in two pharmacies in the Canary Islands, Spain, was used [15]. The patients' identification data have been recoded, so no personal data are accessible. The submitted project was approved by the Research Ethics and Animal Welfare Committee of the University of La Laguna (CEIBA2017-0255). The data corresponds to dispensations made to patients over 18 years of age, over 4 years. However, for the analysis, only patients aged 65 years or older who are targeted by the STOPP/START criteria are included.

Selected criteria

The STOPP criteria studied were those that allow the detection of PIP due to a lack of necessary drug when another treatment is present, in order to prevent the adverse effect or contribute to its therapeutic effect [C11, non-steroidal anti-inflammatories drugs (NSAID) with another antiplatelet agent without prophylactic proton pump inhibitor (PPI) treatment; H8, NSAID with corticosteroids without PPIs at prophylactic dose; L2, use of regular (not on-demand) opioids without concomitant laxative; and L3, long-acting opioids without fast-acting opioids for breakthrough pain], and those due to the inappropriate combination of drugs [A3, any duplicate drug class prescription; B3, beta-blockers in combination with verapamil or diltiazem; B12, aldosterone antagonists with concurrent potassium-conserving drugs without monitoring of serum potassium; B13, phosphodiesterase type-5 inhibitors in severe heart failure characterised by hypotension, or concurrent daily nitrate therapy for angina (risk of cardiovascular collapse); C10, NSAID in combination with vitamin- K antagonist, direct thrombin inhibitor or factor Xa inhibitor; D7, anticholinergics/antimuscarinics to treat extrapyramidal side effects of neuroleptics drugs; D11, acetylcholinesterase inhibitors with a history of persistent bradycardia, heart block or recurrent syncope of unexplained aetiology or concurrent treatment with drugs that lower the heart rate; and N1, concomitant use of 2 or more drugs with antimuscarinic/anticholinergic properties]. In the first case, criterion L2 was selected as the model for applying the method of analysis, and for the second group, criterion B12. In addition, D5 [Benzodiazepines (BZD) for ≥ 4 weeks] was also analysed. PIPs occur when situations such as those described in each of the criteria occur. It should also be noted that as the pharmacies did not have the patient's medical history data, the rest of the STOPP criteria and the START criteria could not be analysed.

Data

Patient electronic prescriptions have been downloaded as a single table database in a MS Excel® sheet following the prescription system implemented by the Canary Health Service. The tables, one for each pharmacy, contain 23 fields with demographic, prescription, and dispensing data [15]. To work in RStudio® (v. 1.2.1335), both were joined into a single file called data.

On the other hand, the Prescription Nomenclature (Nomenclator) of the Spanish Agency for Medicines and Health Products (Spanish acronym, AEMPS) was used, which consists of a database that includes everything related to the identification and technical information of drugs [16]. It contains the Prescription.xml file that collects all the data for each medication. Of all of them, the national code for the presentation of the drug and the DCSA code (dcsaCod) have been selected, which defines the drug from its active principles based on the Snomed-CT terminology (Systematized NOMenclature of MEDIcine – Clinical Terms), regardless of pharmaceutical form or dose [16]. The Nomenclator also contains the DICTIONARY_DCSA.xml file, which contains the dcsaCod and DCSA names (dcsaName) of the drugs. These files are in.xml format.

From the DICTIONARY_DCSA.xml file, all the DCSA codes and denominations were imported into an Access Database® file. All drugs in the DICTIONARY_DCSA.xml file had their ATC code added and another field was added to allow the entry of the STOPP criterion that could be infringed. If a drug is not involved in any of the criteria analysed, the algorithm designed allows the field to be left blank, as well as adding several criteria if the drug is involved in more than one. Finally, the data file was updated in RStudio® so that it now contains all the information from the electronic prescription plus the dcsaName and dcsaCod associated with the national codes of the medicines dispensed, obtained from DICTIONARY_DCSA.xml and Prescription.xml. In this way, a single data file is available for any query on the sample.

It is also important to note that different databases from any country can be used, as long as the necessary conversions for RStudio® are carried out in these databases.

Analysis method

To analyse the STOPP criteria, all the information was unified. To do this, all the drugs with their corresponding STOPP criterion that were in the crit_all table were included in the data file, which contained the information on all the dispensations. Now, this data file collects the STOPP criteria in which each drug of the dispensations is present [see Additional file 1, Picture 1,2]. In case the drug is not related to a certain criterion, "NA" is specified. Table 1 shows four dispensations with their corresponding ATC group and the criteria in which they are present.

Table 1 Sample data file
Full size table

The number of drugs in the same treatment plan was obtained [see Additional file 1, Picture 3, script 6] in RStudio®. In this way, it will be possible to obtain the drugs for each of the treatment plans for each patient and identify whether they are polymedicated, i.e., whether they are prescribed 5 or more drugs. Also, the script that allows grouping and obtaining the number of different doctors who have treated and prescribed medication to the same patient is executed in RStudio® [see Additional file 1, Picture 3, script 7]. This will allow further analysis to determine whether this factor influences the number of PIPs related to the STOPP criteria.

From the data file, all information about the total number of study participants, the number of treatment plans and the subgroups according to age less than or greater than or equal to 65 years is obtained [see Additional file 1, Picture 4, scripts 8–10].

For the analysis of binary outcomes, logistic regression was used, as it allows us to functionally relate a dichotomous variable to a set of independent variables, used, and p < 0.05 was the threshold of significance.

The interpretation of the effects of the categorical variables on the dependent variable was done through the odds ratios, OR, with 95% confidence intervals, using the R function glm(). The variables studied were sex, age, duration of treatment, number of drugs and number of doctors for each treatment plan. We transformed age into a categorical variable, so that if at the start of the first treatment plan, the patient was less than 65 years old, a 0 is assigned in the Age column, and if the patient was 65 years old or older, a 1 is assigned. The number of drugs has also been transformed into a categorical variable, i.e., if there are less than 5 drugs in a treatment plan, a 0 is assigned, and if there are 5 or more drugs prescribed it takes the value 1.

The dependent variable, the criterion being analysed, takes values of 1 if the infringement is detected, or 0 when there is no infringement. The control group is made up of individuals whose prescriptions do not violate the criterion under study.

Application of the method to STOPP criteria

As previously mentioned, the L2 and B12 criteria were taken as examples to explain the implementation of the developed method. It should be noted that depending on the type of criterion, the method varies slightly, so that two groups can be distinguished:

  1. a)

    Criteria for which the method of analysis uses a scheme we have termed "without", corresponding to STOPP criteria that are violated by the absence of a drug necessary to prevent the adverse effect of another drug or to contribute to its therapeutic effect, such as criterion L2.

  2. b)

    Criteria for which a so-called "with" scheme is used, i.e., STOPP criteria to detect PIP due to inappropriate drug combinations, as is the case for criterion B12.

a) Application of the "without" method to criterion L2

A infringement of criterion L2 occurs when opioids are used chronically without laxatives (risk of severe constipation). That is, when opioids (ATC group N02A) and drugs used for addictive disorders (ATC group N07B) are prescribed in the same treatment plan, without laxatives (ATC group A06A). In the data file, the script column is filtered to obtain the dispensing lines where the L2 criterion is present, grouped by treatment plan [see Additional file 1, Picture 5, script 11].

The table obtained is renamed S1.L2 and with this developed script it is sought that, in the same dispensation, one of the ATC groups representing opioids is present, p1 ≠ 0, and that the ATC group A06A representing laxatives is not present (p3 = 0), where p is the set of ATC groups involved in the criterion. When these two conditions are met, i.e., p1 ≥ 1 and p3 = 0, a breach of criterion L2 is detected. In addition, the corresponding patient's NUSS (anonymised code that is assigned to each patient), duration of treatment, gender and age, prescribing physician, number of drugs for the same treatment plan and the number of total physicians who have treated the patient are included [see Additional file 1, Picture 5, script 11]. Table 2 shows some treatment plans from Table S1.L2 that have an opioid, p1 ≥ 1, no laxatives, p3 = 0, thus L2 = 1. Also, a case is shown where p1 ≥ 1 and p3 > 1, thus L2 = 0.

Table 2 Sample of various treatment plans with infringement of criterion L2
Full size table

Subsequently, the patient data were grouped according to their NUSS [see Additional file 1, Picture 5, script 12], obtaining the total number of treatment plans for each patient and how many of them violate the L2 criterion, as well as demographic data such as age and sex; this way, Table S2.L2 is obtained. A sample of 7 patients is shown in Table 3, where it is observed that, except for two of them, NUSS 504D288480302 and 24C48C4560317, the L2 criterion is violated in all their treatment plans.

Table 3 Patients where criterion L2 is infringed
Full size table

The logistic regression of the L2 criterion was performed [see Additional file 1, Picture 6, script 13] as well as the graphical representations of the results [see Additional file 1, Picture 7, scripts 14 and 15]. The dependent variable L2 takes values of 1, if the infringement is detected, or 0 when there is no infringement. The control group, L2 = 0, consists of individuals who have been prescribed opioids as well as laxatives.

b) Application of the "with" method to criterion B12

Infringement of criterion B12 occurs if an aldosterone antagonist is prescribed together with other drugs that may increase potassium levels without monitoring potassium (risk of severe hyperkalaemia > 6.0 mmol/L; serum potassium should be monitored regularly, at least every 6 months). Therefore, for the infringement to be identified, active principles belonging to ATC groups C03D or C03E including aldosterone antagonists and ATC groups C09A, C09B, C09C, C09D, C10B or C07D including all drugs that increase potassium levels: angiotensin-converting enzyme inhibitors (ACE inhibitors), angiotensin II receptor blockers (ARB), amiloride, triamterene, must be present in the same dispensing. In the data file, the script column is filtered to obtain all the dispensing lines where the B12 criterion is present, grouped by treatment plan [see Additional file 1, Picture 8, script 16]. The table obtained is renamed S1.B12 and the script developed seeks to ensure that one of the ATC groups representing aldosterone antagonists, p1 ≠ 0, combined with one of the ATC groups representing drugs that increase potassium levels, p2 ≠ 0, is present in the same dispensation. When these two conditions are met, i.e., p1 ≥ 1 and p2 ≥ 1, a breach of criterion B12 is detected. In addition, once the table with the treatment plans is obtained, the corresponding patient's NUSS, duration of treatment, gender and age, prescribing physician, number of drugs for the same treatment plan and the number of total physicians who have treated the patient are included [see Additional file 1, Picture 8, script 16].

Table 4 shows a sample of Table S1.B12, with several treatment plans where criterion B12 is violated. As can be seen, for example in treatment plan 10011340972958, columns p1 and p2 take the value 2, which means that an active substance from group C03D or C03E has been dispensed twice in the case of p1 and in the case of p2 from group C09A, C09B, C09C, C09D, C10B or C07D, thus infringing criterion B12.

Table 4 Sample of various treatment plans with infringement of criterion B12
Full size table

To obtain the total number of treatment plans for each patient and how many of them violate criterion B12, the data are grouped according to their NUSS [see Additional file 1, Picture 8, script 17]. This also allows demographic data such as age and gender to be obtained, resulting in Table S2.B12. Table 5 shows a sample of 7 patients from Table S2.B12 with infringement of STOPP criterion B12; for example, patient NUSS 537C3A3340429 has 28 dispensed treatment plans of which 23 are in infringement of STOPP criterion B12.

Table 5 Patients where criterion B12 is infringed
Full size table

For B12 criterion, logistic regression was also performed [see Additional file 1, Picture 9, script 18]. The dependent variable B12 can take the values 1 if a infringement of criterion B12 is detected or 0 when there is no infringement. The control group, B12 = 0, is formed by selecting individuals who have been prescribed drugs from the ATC groups that may lead to a breach of criterion B12, but who are not prescribed aldosterone antagonists, so no breach occurs.

Therefore, the method is able to detect infringements of the criteria for both the without method and the with method.

A separate case is criterion D5 (BZD for ≥ 4 weeks), so in order to detect a possible infringement of this criterion in medical prescriptions, a slight modification of the analysis method was applied from the data file. Pictures 10 and 11 (see Additional file 1) show the scripts to obtain the number of total patients and those patients ≥ 65 years old taking BZD for ≥ 4 weeks. First, from the drugs file (contains active principles that increase the risk of falls, with their pharmacological groups, ATC group, dcsaCod and dcsaName) [17] the pharmacological group is added to the data file and then all dispensing lines that do not include BZD are deleted; Finally, patients are grouped, and the difference between the end date of the treatment plan and the start date of the BZD package is obtained. If the difference between these is ≥ 28 days, a breach of criterion D5 is identified.

Results

A sample of 10 460 patients was obtained, 2 312 of whom were equal to or older than 65 years of aged (22.1%) (Table 6) and of these, 56% were women. Almost half of the patients ≥ 65 years (46.6%) presented polypharmacy (≥ 5 drugs); 9.3% of them had excessive polypharmacy (≥ 10 drugs). On the other hand, 70% of the total patients were treated by one doctor, 19% by two doctors and 11% of the patients by more than two doctors.

Table 6 Patients and treatment plans presenting the following STOPP criteria
Full size table

In the total sample, at least one PIP was detected in 39.3% of patients and in 34.9% of their treatment plans. Among the patients, the most detected infringement corresponds to criterion D5 (20.8%), followed by L2 (9.5%) and A3 (3.9%) (Table 6).

In the group of patients ≥ 65 years 57% PIPs was detected on patients and in 38% of their treatment plans (Table 6). In this age group infraction criterion D5 was detected in 25.9% of patients, L2 in 13.8% and A3 in 5.7% (Table 6). From 131 patients in which the infraction of criterion A3 was detected, the most numerous duplicities were BZD (39%) and NSAID (37. 4%).

Multiple infractions were detected in 2.3% of the total patients and 2.1% in their treatment plans, as well as 5.1% of the patients ≥ 65 years and 2.7% of their treatment plans. The most common multiple infractions were those of criteria L2 and L3.

Focusing on the two criteria used as models for the development of the analysis method, the results were as follows: criterion L2 is violated in 13.8% of patients ≥ 65 years and 7.4% of their treatment plans, representing 3.1% and 2.7% of the total, respectively (Table 6). Criterion B12 is not met in 1.8% of patients ≥ 65 years and 1.2% of the treatment plans of this group, which represents approximately 4 ‰ of the total number of patients and treatment plans (Table 6).

The data for the rest of the criteria can be found in Table 6, except for criterion B13, of which no case was detected, because during the study period, phosphodiesterase 5 inhibitors were not prescribed by electronic prescription. However, currently said prescription is included in the electronic prescription, with which, if the infraction were to occur, the corresponding dispensation would be detected.

Table 7 shows the OR values for each variable with their 95% CI for STOPP criteria studied. Logistic regression could not be applied in the cases of criterion D11 due to the small number of patients and in criteria A3 and D5 due to the characteristics of their analysis.

Table 7 OR values and 95% CI for the variables of the STOPP criteria
Full size table

Regarding the variables studied, in the case of age, the null hypothesis is rejected for criteria B12, D7, L2, L3 and N1. In the case of criteria B12 and L3, being over 65 years of age implies a greater risk of suffering a infringement of the criteria, while for criteria D7, L2 and N1, when calculating the inverse of the OR to better express the strength of the relationship, it is observed that patients under 65 years of age have a risk of 2.6, 2.2 and 3 times more, respectively, of being prescribed a medication that violates said criteria (Table 7).

For the sex variable, only in the case of the L2 criterion, the fact of being a woman implies a greater risk of suffering this PIP. On the contrary, being a man increases the risk of presenting PIPs according to criteria B3, C10, D7 and N1 (Table 7).

Regarding the number of drugs, polymedicated patients have a higher risk of having PIPs related to criteria B3, B12, C10, C11, D7, L3 and N1; on the contrary, in non-polymedicated patients there is a risk of approximately 3 times compared to polymedicated patients of infringing the L2 criterion (Table 7).

Patients treated with several physicians have a higher risk of suffering PIPs related to criteria D7, L2, L3 and N1 (Table 7).

The duration of treatment was only significant for the criteria L2 and H8, increasing or decreasing, respectively, the probability of the infringement occurring with increasing duration of treatment (Table 7).

The representation of the results was made for the significant variables of the criteria studied and as a sample those corresponding to the L2 criterion are exposed. Figure 1 shows the probability for the variables, age, sex, number of drugs and number of doctors. For sex, the populations are overlapping and the medians and 95th percentile are very close. For age, the populations are overlapping, and there is a difference between the median of patients aged 65 and over, with a value of around 0.85, and the median of younger patients, with 0.95. Regarding the number of drugs, we observe that the populations overlap and the probability of the L2 criterion occurring is lower for patients with five or more drugs (p = 0.85), compared to those with no polypharmacy (p = 0.90). Finally, for the variable nPhysicians, the higher the number of doctors, the higher the probability of the L2 criterion being violated.

Fig. 1
figure 1

Relationship between probability and variables studied for criterion L2

Full size image

This fact is better appreciated in Fig. 2, where for both sexes, age 65 years or over, and polymedicated, the risk is higher in those patients with a greater number of physicians. Also, it can be observed that, regardless of the number of doctors, the risk increases slightly as the duration of treatment increases and that the probability of the L2 criterion occurring is greater in women ≥ 65 years polymedicated than in men in the same conditions.

Fig. 2
figure 2

Relationship between probability and duration of treatment as a function of the number of doctors for criterion L2

Full size image

However, if we look at the group of women under 65 years of age and not polymedicated, we observe that the probability of the L2 criterion being violated is independent of the duration of treatment and the number of physicians who have prescribed to this group. The probability remains practically constant throughout the duration of treatment for both one doctor and nine. In polymedicated women under 65, the probability decreases from practically 1 in the case of nine doctors, to 0.8, in the case of a single doctor. However, both values get closer as the duration of treatment increases, the same behavior as in the case of polymedicated women over 65 years of age, although in these, the differences between 1 and 9 doctors are much more pronounced.

Discussion

Polypharmacy is one of the factors related to PIPs. The published prevalence data are highly variable and show a progressive increase over the years [2, 9, 18, 19]; In the present study, carried out with medical prescriptions from two community pharmacies, 46.6% of patients ≥ 65 years of age are polymedicated (use of 5 or more drugs), a value higher than that obtained for the Spanish population ≥ 65 years of age, 28.2% [2], and within the range of 27%—59% published in a systematic review of studies conducted under similar conditions [20]. In the case of outpatients and institutionalized in social health centers, combination therapy ranges from 30–60% [21, 22], even reaching 79.8%, but in patients ≥ 70 years [23]. This variability in the data may be affected by the degree of pharmacotherapeutic follow-up, or by the accessibility of the family doctor and the health center. If we consider prescriptions of 6 or more drugs as polypharmacy, the value ranges between 21.7% and 67% [13, 24, 25], reaching 93.8% in hospitalized patients [9].

In our group of polymedicated patients ≥ 65 years, 9.3% have been prescribed more than 10 drugs, compared to the 3.4% obtained for the Spanish population ≥ 65 years [2], or between 2% and 20.2% in patients institutionalized [23, 24]. In hospitalized patients, the values range between 15 and 59% [9, 26]. Other authors publish similar data but broken down by age intervals, polypharmacy and hyperpolypharmacy increasing with age [2, 27]. This increase may be due to the established habit that the greater the number of diseases and complications derived from them, the greater the number of drugs prescribed, without carrying out a periodic review of the prescribed drugs and the needs of the patient. Many times, it is not analyzed whether the time necessary to achieve the benefit of the medication will be greater than the patient's life expectancy.

Regarding the application of the STOPP/START criteria, we have previously commented on the impossibility of analyzing all of them due to the lack of access to the patient's clinical data. In fact, as indicated by Carvalho et al. (2019) [28], if only the list of medications taken by the patient is available, only a third of all the STOPP criteria and only one START criteria can be analyzed. Therefore, it is necessary to know the diagnostic and laboratory data of the patient.

The results for each of the criteria analyzed are discussed below, considering that PIPs were detected in 57% of patients ≥ 65 years of age and in 38% of their treatment plans.

A3 criterion

This criterion is infringed in 5.7% of patients and in 2% of their treatment plans. In several studies carried out in primary care, drug duplication was the most frequently identified STOPP criteria [29, 30], approaching or even exceeding 10% in some cases [31,32,33]. Although most of the authors do not specify which duplications of drugs they have considered, some cite duplications of opioids, NSAID, selective serotonin reuptake inhibitors, loop diuretics, and ACE inhibitors [31]. In our case, the most frequent correspond to duplications of NSAID and duplications of BZD. Two NSAID are among the top twenty most widely consumed active ingredients in number of packages in Spain: ibuprofen and dexketoprofen, and several BZD [34]. It is not surprising, therefore, that for criterion A3, duplication of these drugs constitutes the most frequent infringements, both in the total population and in patients ≥ 65 years of age. For intervals between 65 and 69 years, the prevalence is 8.5% [24] and in patients ≥ 70 years of primary care, the frequency ranges from 2.4% to 6.3% [10, 11, 13, 35]. The origin of the data may contribute to this variability, which, although all correspond to out-of-hospital patients, these may be outpatients, residents of social health centers, or both at the same time.

At the hospital level, the studies carried out normally involve a certain area, in addition to the fact that the data can come from admission to the emergency room, hospitalization or even hospital discharge, so the disparity can be important, finding values between 4% and 14.5% [9, 36,37,38,39].

Many of the cited works applied the first version of the STOPP/START criteria [40], while others the second [7]. This fact, however, does not justify the differences observed in the prevalence of criterion A3, since between both versions there were no changes regarding this criterion. However, other variables such as the method of analysis, the age of the patients, or the duration of the study, may have contributed to the disparity of results observed. Although our work was carried out with data from electronic dispensing in community pharmacies over a period of 4 years, in all the cited cases it is found that drug duplication is a common practice both in primary care and in the hospital setting, which poses a risk to the patient.

B3 criterion

Our findings indicate that the percentage of patients and treatment plans in people over 65 years of age is very low (Table 6), unlike other higher values, perhaps due to the size of the sample used, age, or the level of care in which the study was carried out, hospital or community [13, 28, 30, 41, 42]. Other studies carried out in primary care such as ours indicate a very similar percentage of patients with beta-blocker prescriptions associated with verapamil, 0.2% [10, 11], but with the difference that they only include patients from 70 years old and they only consider verapamil. In addition, they indicate that women are at greater risk of suffering this PIP, unlike in our case, where the risk is 4 times higher for men (Table 7), even though the percentage of women in both samples is similar. This may be due to differences in the clinical protocols for action in cardiac diseases or to a higher prevalence of these diseases in men in our country. On the other hand, polymedicated patients have a risk 10 times higher than non-polymedicated patients of suffering this PIP.

B12 criterion

This PIP is identified by other authors in 3—5% of prescriptions [37, 42], while in our work we detected 1.2% (Table 6). Regarding patients, published prevalence data ranges from 2.3% to 12.5% [9, 13, 24, 42], our result being lower, 1.8%. All this denotes a great variability in the prevalence of this PIP, due to the sample size, age group of patients, method used to detect PIP, and level of care, hospital, or extra-hospital.

The coincidence of both groups of drugs in the same prescription, without knowing the potassium levels in the blood, does not imply a infringement of criterion B12. For this reason, we cannot affirm that the patients will suffer hyperkalemia; however, we see that if potassium levels are not monitored, there is a risk of hyperkalemia due to the use of these drugs [9, 37, 42]. If we knew the analytical data, we could include it in the RStudio® together with the dispensing data and thus quickly detect patients with hyperkalemia or risk of suffering it and who are taking aldosterone antagonist drugs with other drugs that can increase the levels of potassium.

C10 criterion

In our sample, the criterion is violated in 0.43% of the patients, similar to other published data [9, 39, 41, 42]; however, it contrasts with the results obtained by other authors, with higher values [13, 28]. This PIP was detected in 0.14% of the treatment plans, a much lower value than others published [37].

C11 criterion

In the sample of patients aged ≥ 65 years, 2.77% present this PIP, while in other studies the values range from 0.1% to 14.6% [13, 24, 28, 39, 41], although with different age ranges and levels of patient care. In our case, it is not surprising to find few patients with this PIP, given that the use of Proton Pump Inhibitor (PPI) in Spain is very high and has been increasing over the years; thus, in 2021, global consumption (Defined Daily Dose per 1000 Inhabitants and days, DHD) was 130.7, corresponding to omeprazole 78.1% [43]. This high consumption has also been revealed in other countries such as Ireland [11].

D5 criterion

According to the Annual Report of the National Health System for the year 2017, four BZD (lorazepam, alprazolam, lormetazepam and diazepam) were among the 20 active ingredients with the highest number of invoiced containers, from which the high consumption of these drugs in Spain can be deduced [34]. Numerous publications from both hospitals and out-of-hospital settings confirm this, showing that 4.4% to 7.6% of patients are prescribed BZD with a long half-life lasting longer than a month. This is one of the most common infringements of STOPP criteria at any level of care [29, 31, 35, 39, 44]. In our case, considering the total number of patients ≥ 65 years (n = 2312), 29% have been prescribed some BZD and of these, approximately 26% have it for 28 days or more, constituting a PIP according to the D5 criterion. From these results it can be deduced that the most of patients who start treatment with a BZD continue with it after 28 days. Of these patients, 68% are women. Likewise, the infringement of criterion D5 is detected in 21.5% of their treatment plans. It must be taken into account that we have considered all the BZD and that the data correspond to four years, which may have influenced the result, although they are similar to other published [9, 13, 23, 36, 38, 41, 42, 45].

For all these reasons, it can be concluded that BZD are highly prescribed among the elderly both in Spain and in other countries and their use continues to increase, finding that their prescription is well above what would be appropriate, also considering, that are associated with negative outcomes such as dependency, falls and fractures, cognitive impairment, or sleep disturbances.

D7 criterion

This PIP was detected in very few cases, only in 0.13% of patients ≥ 65 years of age and of their treatment plans, data consistent with other published data [9,10,11, 13, 35, 41], despite the fact that many of these works do not specify the prescribed neuroleptic, nor the anticholinergic and antimuscarinic drugs that have been evaluated, the age groups do not coincide; In addition, patients may or may not be institutionalized, and some even record the infringement when these anticholinergics are prescribed for at least four weeks [41], while, in the present study, as long as there is a prescription for a neuroleptic together with an anticholinergic in the treatment plan, RStudio® detects it. With the analysis method presented in RStudio® and validated in this sample, we can identify those patients and treatment plans with anticholinergics or antimuscarinics and neuroleptics where there is a risk of anticholinergic toxicity. We can carry out a review of the pharmacological treatment, minimising the possible adverse effects that could arise.

Taking all this into account, we can conclude that the frequency of PIPs related to criterion D7 is very low in the population ≥ 65 years of age and that, according to our results, being a man, being under 65 years of age, being on polypharmacy or having been treated for various physicians, poses a higher risk of having this PIP. Other authors reach the same conclusion regarding age [10, 11], but nevertheless, with respect to sex, the studies are contradictory.

D11 criterion

We have detected prescriptions of acetylcholinesterase inhibitors with drugs that decrease heart rate in 0.3% of patients, a value like that of other authors [9, 39], but much lower than other published values [13, 28, 41]. In the case of this criterion, all the infringements detected in the total sample coincide with patients ≥ 65 years of age (Table 6).

H8 criterion

1.34% of patients ≥ 65 years of age have this PIP, which would imply an increased risk of peptic ulcer. This data is also similar to that obtained by other authors [13, 28, 41, 45], despite the different characteristics in terms of age and the level of care the patient is in. It may still be that the frequency of this PIP is less, since many patients acquire the PPI without a medical prescription, with which there would be no infringement of the criterion; However, our method detects it because we only included data from the treatment plan. Likewise, an association was found between the duration of treatment and the probability of suffering from this PIP, in such a way that as the former increases, the probability decreases before the infraction occurs (Table 7). This fact is logical since it is probable that, as time passes, the treatment will be reviewed, and the absence of the PPI will be detected.

L2 and L3 criteria

These STOPP criteria involve opiate-related PIPs. In our study, 7.4% of prescribed opioid prescriptions without laxatives (criterion L2) were detected in almost 14% of patients ≥ 65 years of age (Table 6), with the consequent risk of constipation. In this case, not only is there a greater disparity in the prevalence of prescriptions [37, 45, 46] and of patients with this PIP [9, 28, 38, 39, 41], but also lower than that obtained by us, which is surprising considering the frequent use of opioids in many countries. In Spain, for example, tramadol (associated with paracetamol) and fentanyl are among the top twenty most consumed active ingredients [47]. Its presence, therefore, in our treatment plans, is quite common. However, here we also find differences in the characteristics of the patient sample, in addition to the fact that the disparity observed in the different publications could be due to the different interpretation of the criterion by the authors, although most do not specify anything in this regard. Some consider that the L2 criterion is violated when no laxatives have been prescribed after six months of opioid treatment [41], while our method identifies it from the moment the prescription includes an opioid prescribed without a laxative.

Infringement of the L3 criterion was detected in 3.6% of patients ≥ 65 years and in 2.1% of their treatment plans. Once again, the prevalence of this PIP is higher than that reported in other publications [39, 41]. As in the case of L2, the explanation could be the interpretation of the criterion; thus, some consider that it is violated when short-acting opioids have not been prescribed after one month of treatment with long-acting opioids [41], while the proposed method of analysis detects the infringement from the moment a long-acting opioid is prescribed in a treatment plan without a short-acting opioid for breakthrough pain.

The risk of infringing the L2 criterion is double in women compared to men, coinciding with other researchers [38], while in the older age group there is a lower probability of the infringement occurring, contrary to L3 criterion. Polymedicated patients are at higher risk of infringing L2 and L3 criteria. Lastly, the probability that the L2 and L3 criteria will be violated increases with the number of physicians and, in the case of L2, with the duration of treatment.

The proposed and validated method allows us to carry out a more in-depth analysis of the group of patients who are taking opioids, both for cancer and non-cancer pain, and who have not been prescribed a laxative to prevent constipation. In addition, it allows us to identify patients with prescribed opioids without a rescue regimen for acute pain, or that this regimen is insufficient in case of uncontrollable pain. This makes it possible to monitor the population under treatment and locate that group with difficult pain management.

N1 criterion

In our sample of patients ≥ 65 years, 1.08% concomitantly receive two or more drugs with anticholinergic or antimuscarinic properties, while the published data range between 0.17 and 2.9% [9, 28, 41]. PIP is detected in 2.26% of their treatment plans. The pharmacological groups considered for the analysis of this criterion were, among others, bladder and intestinal antispasmodics, tricyclic antidepressants, and first-generation antihistamines, coinciding with other authors [41]. Lastly, with this method of analysis we can quickly identify treatment plans and patients at risk, identifying possible interactions, adverse effects, or cascading prescriptions, so frequent in these groups of drugs with anticholinergic or antimuscarinic properties.

Strengths and limitations

The strengths of this study include the ease that the method developed presents for its application in a large group of patients in a short period of time, regardless of the origin of the data and the level of care in which the patient is, and for any National System of Health. The algorithm allows us to redefine any concept and filter by variables such as age, sex, or any other characteristic for which we want to obtain information from the patients. It is a useful tool to assess the effectiveness of a medical or pharmaceutical intervention on the treatment of a patient. For example, we can find out if a program being carried out to reduce polypharmacy or to detect PIPs is being effective. To do this, it is enough to filter the data by dispensing date before and after the intervention program and finally, analyze the data.

With this method, we could analyze other variables based on the diagnosis, if we had that information, and we could even identify patients at risk of BZD addiction and, in this way, prevent and carry out pharmacotherapeutic follow-up.

In summary, this method has been validated in a large sample of outpatients, and at the care level it is a great tool due to the speed and flexibility it presents when it comes to obtaining information.

The study has limitations because it has not been possible to analyse all the START criteria and the rest of the STOPP criteria, although the proposed method allows this to be done. The reason is that Spanish community pharmacists do not have access to patients' medical records.

Conclusions

The method developed and validated in RStudio® with different STOPP criteria, makes it possible to analyze the pharmacological treatment of many patients with different characteristics and to identify those at risk of suffering some PIP according to the STOPP criteria, in order to prevent side effects. The results of the study confirm that there is a high prevalence of PIPs in patients ≥ 65 years of age. The most frequent PIPs associated with the STOPP criteria were the use of BZD for more than 28 days, the use of prescribed opioids without associating a laxative, and duplication of drugs. The most numerous duplications were from BZD and NSAID. Of the multiple infractions detected in the same patient, the most frequent corresponds to the L2 and L3 criteria, referring to opioids. Our results indicate that polypharmacy is the most common risk factor affecting PIPs for most criteria, being present in almost half of patients ≥ 65 years of age.

From the validated criteria, it appears that the risk of infringement of criteria D7, L2, L3 and N1 increases with the number of doctors prescribing treatment to the patient. For the other criteria, their influence is not significant.

This study highlights the need for medication review by pharmacists and their presence in multidisciplinary geriatric teams, both at the hospital and community level, to promote the rational use of medication and improve quality of life of these. The pharmacotherapeutic follow-up of the patient may imply the need to deprescribe the medication.

Availability of data and materials

The datasets generated and/or analysed during this study are not publicly available because they are confidential data belonging to the Canary Islands Health Service (SCS) of the Canary Islands Government. For the same reason, they are also not available from the corresponding author.

Abbreviations

ACE inhibitors:

Angiotensin-converting enzyme inhibitors

AEMPS:

Spanish Agency for Medicines and Health Products

ARB:

Angiotensin II receptor blockers

BZD:

Benzodiazepines.

DCSA code:

DcsaCod

DCSA names:

DcsaName

Nomenclator:

Prescription Nomenclature of the AEMPS

NSAID:

Non-steroidal anti-inflammatories drugs.

NUSS:

Anonymised code that is assigned to each patient

PIP:

Potentially inappropriate prescribing.

PPI:

Proton pump inhibitor.

STOPP/START:

Screening Tool of Older Person’s potentially inappropriate Prescriptions / Screening Tool to Alert doctors to Right Treatment

References

  1. Romero T. Number of inhabitants aged 65 or over in spain from 2002 to 2022 [Web document]. , 2022 [cited Dec 15, 2022]. <https://www.statista.com/statistics/1231682/population-of-spain-over-65-year/>.

  2. Hernández-Rodríguez MÁ, Sempere-Verdú E, Vicens-Caldentey C, González-Rubio F, Miguel-García F, Palop-Larrea V, Orueta-Sánchez R, Esteban-Jiménez Ó, Sempere-Manuel M, Arroyo-Aniés MP, Fernández-San JB. Evolution of polypharmacy in a spanish population (2005–2015): A database study. Pharmacoepidemiol Drug Saf. 2020;29(4):433–43. https://doi.org/10.1002/pds.4956.

    Article  PubMed  Google Scholar 

  3. Martin-Pérez M, López de Andrés A, Hernández-Barrera V, Jiménez-García R, Jiménez-Trujillo I, Palacios-Ceña D, Carrasco-Garrido P. Prevalencia de polifarmacia en la población mayor de 65 años en españa: Análisis de las encuestas nacionales de salud 2006 y 2011/12. Revista española de geriatría y gerontología 2017;52(1):2–8. https://doi.org/10.1016/j.regg.2016.07.006.

  4. Delgado Silveira E, Montero Errasquín B, Muñoz García M, Vélez-Díaz-Pallarés M, Lozano Montoya I, Sánchez-Castellano C, Cruz-Jentoft AJ. Mejorando la prescripción de medicamentos en las personas mayores: Una nueva edición de los criterios STOPP-START. [improving drug prescribing in the elderly: A new edition of STOPP/START criteria]. Rev Esp Geriatr Gerontol. 2015;50(2):89–96. https://doi.org/10.1016/j.regg.2014.10.005.

  5. Page RL II, Linnebur SA, Bryant LL, Ruscin JM. Inappropriate prescribing in the hospitalized elderly patient: Defining the problem, evaluation tools, and possible solutions. Clin Interv Aging. 2010;1(5):75–87. https://doi.org/10.2147/CIA.S9564.

    Article  Google Scholar 

  6. O’Sullivan DP, O’Mahony D, Parsons C, Hughes C, Murphy K, Patterson S, Byrne S. A prevalence study of potentially inappropriate prescribing in irish long-term care residents. Drugs Aging. 2013;30(1):39–49. https://doi.org/10.1007/s40266-012-0039-7.

    Article  PubMed  Google Scholar 

  7. O’Mahony D, O’Sullivan D, Byrne S, O’Connor MN, Ryan C, Gallagher P. STOPP/START criteria for potentially inappropriate prescribing in older people: Version 2. Age Ageing. 2015;44(2):213–8. https://doi.org/10.1093/ageing/afu145.

    Article  PubMed  Google Scholar 

  8. O’Mahony D, Cherubini A, Guiteras AR, Denkinger M, Beuscart J, Onder G, Gudmundsson A, Cruz-Jentoft AJ, Knol W, Bahat G, van der Velde N, Petrovic M, Curtin D. STOPP/START criteria for potentially inappropriate prescribing in older people: Version 3. Eur Geriatr Med. 2023. https://doi.org/10.1007/s41999-023-00777-y.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Baré M, Lleal M, Ortonobes S, Gorgas MQ, Sevilla-Sánchez D, Carballo N, De Jaime E, Herranz S. Factors associated to potentially inappropriate prescribing in older patients according to STOPP/START criteria: MoPIM multicentre cohort study. BMC Geriatr 2022 -01–11;22(1). https://doi.org/10.1186/s12877-021-02715-8.

  10. Cahir C, Fahey T, Teeling M, Teljeur C, Feely J, Bennett K. Potentially inappropriate prescribing and cost outcomes for older people: A national population study. Br J Clin Pharmacol. 2010;69(5):543–52. https://doi.org/10.1111/j.1365-2125.2010.03628.x.

    Article  PubMed  PubMed Central  Google Scholar 

  11. Bradley MC, Fahey T, Cahir C, Bennett K, O’Reilly D, Parsons C, Hughes CM. Potentially inappropriate prescribing and cost outcomes for older people: A cross-sectional study using the northern ireland enhanced prescribing database. Eur J Clin Pharmacol. 2012;68(10):1425–33. https://doi.org/10.1007/s00228-012-1249-y.

    Article  PubMed  Google Scholar 

  12. Díaz-González A, Morera-Herreras T, Medibil-Crespo I, Pisón-Rodríguez J, Martín-Lesende I. Prescripción inadecuada, crónica y a demanda, en pacientes pluripatológicos ≥65 años según los criterios STOPP/START. Gac Med Bilbao. 2018;115(2):45–57.

    Google Scholar 

  13. Fuertes Abardía C, Ballesta Juan P, Cruz Esteve I, Galindo Ortego G, Marsal Mora JR, Gómez-Arbonés X. Prescripción potencialmente inadecuada: Utilidad de los criterios STOPP/START versión 2 a nivel poblacional en cataluña. SEMERGEN. 2022;48(3):163–73. https://doi.org/10.1016/j.semerg.2021.09.005.

    Article  PubMed  Google Scholar 

  14. Instituto Canario de Estadística (. Demographic data on the canary islands [Web document]. , 2022 [cited 03/30/2023]. <https://www.gobiernodecanarias.org/istac/estadisticas/>.

  15. González Munguía S. Detección de errores de medicación en pacientes polimedicados. Tesis Doctoral. Universidad de La Laguna. http://riull.ull.es/xmlui/handle/915/25572, 2020.

  16. Ministerio de Sanidad, Servicios Sociales e Igualdad. Agencia Española de Medicamentos y Productos Sanitarios. Centro de información online de medicamentos de la AEMPS – CIMA. nomenclátor de prescripción.[Web document]. [cited May 30, 2015]. <https://cima.aemps.es/cima/publico/nomenclator.html>.

  17. González-Munguía S, Munguía-López O, Sánchez Sánchez E. Pharmacist comprehensive review of fall-risk-increasing drugs and polypharmacy in elderly spanish community patients using RStudio. Heliyon 2023;9(6):e17079. https://doi.org/10.1016/j.heliyon.2023.e17079.

  18. Rollason V, Vogt N. Reduction of polypharmacy in the elderly: A systematic review of the role of the pharmacist. Drugs Aging. 2003;20(11):817–32. https://doi.org/10.2165/00002512-200320110-00003.

    Article  PubMed  Google Scholar 

  19. Wastesson JW, Morin L, Tan ECK, Johnell K. An update on the clinical consequences of polypharmacy in older adults: A narrative review. Expert Opin Drug Saf. 2018;17(12):1185–96. https://doi.org/10.1080/14740338.2018.1546841.

    Article  PubMed  Google Scholar 

  20. Elmstahl S, Linder H. Polypharmacy and inappropriate drug use among older people-a systematic review. Healthy Aging & Clinical Care in the Elderly. 2013;16:1–8. https://doi.org/10.4137/HACCE.S11173.

    Article  Google Scholar 

  21. Wastesson JW, PhD, Canudas-Romo V, PhD, Lindahl-Jacobsen R, PhD, Johnell K, PhD. Remaining life expectancy with and without polypharmacy: A register-based study of swedes aged 65 Years and older. J Am Med Dir Assoc. 2016;17(1):31–5. https://doi.org/10.1016/j.jamda.2015.07.015.

  22. Eiras A, Teixeira MA, González-Montalvo JI, Castell M, Queipo R, Otero Á. Consumo de medicamentos en mayores de 65 años en oporto (portugal) y riesgo de prescripción de medicamentos potencialmente inapropiados. Aten Primaria. 2016;48(2):110–20. https://doi.org/10.1016/j.aprim.2015.03.005.

    Article  PubMed  Google Scholar 

  23. Díez R, Cadenas R, Susperregui J, Sahagún AM, Fernández N, García JJ, Sierra M, López C. Potentially inappropriate medication and polypharmacy in nursing home residents: A cross-sectional study. J Clin Med. 2022;11(13):3808. https://doi.org/10.3390/jcm11133808.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Krustev T, Milushewa P, Tachkov K, Mitov K, Petrova G. Evaluation of potentially inappropriate medication in older patients with cardiovascular diseases-STOPP/START-based study. Front Public Health. 2022;22(10):1023171. https://doi.org/10.3389/fpubh.2022.1023171.

    Article  Google Scholar 

  25. Mud Castelló F, Mud Castelló S, Rodríguez Moncho MJ, Ivorra Insa MD, Ferrándiz Manglano ML. Herramientas para evaluar la adecuación de la prescripción en ancianos [tools for evaluating the suitability of prescriptions for the elderly]. Farm Comunitarios. 2013;5(4):147–51.

    Google Scholar 

  26. Fabbietti P, Ruggiero C, Sganga F, Fusco S, Mammarella F, Barbini N, Cassetta L, Onder G, Corsonello A, Lattanzio F, Di Rosa M. Effects of hyperpolypharmacy and potentially inappropriate medications (PIMs) on functional decline in older patients discharged from acute care hospitals. Arch Gerontol Geriatr. 2018;77:158–62. https://doi.org/10.1016/j.archger.2018.05.007.

    Article  PubMed  Google Scholar 

  27. Suzuki Y, Sakakibara M, Shiraishi N, Hirose T, Akishita M, Kuzuya M. Prescription of potentially inappropriate medications to older adults. A nationwide survey at dispensing pharmacies in japan. Arch Gerontol Geriatr. 2018;77:8–12. https://doi.org/10.1016/j.archger.2017.12.010.

  28. Carvalho R, Lavrador M, Cabral AC, Veríssimo MT, Figueiredo IV, Fernandez-Llimos F, Castel-Branco MM. Patients’ clinical information requirements to apply the STOPP/START criteria. Int J Clin Pharm. 2019;41(6):1562–9. https://doi.org/10.1007/s11096-019-00920-5.

    Article  PubMed  Google Scholar 

  29. Salgueiro E, Elizarde BC, Elola AI, García-Pulido B, Nicieza-García ML, Manso G. Los criterios STOPP/START más frecuentes en españa. una revisión de la literatura. Rev Esp Geriatr Gerontol. 2018;53(5):274–8. https://doi.org/10.1016/j.regg.2018.03.001.

  30. Ryan C, O’Mahony D, Kennedy J, Weedle P, Byrne S. Potentially inappropriate prescribing in an irish elderly population in primary care. Br J Clin Pharmacol. 2009;68(6):936–47. https://doi.org/10.1111/j.1365-2125.2009.03531.x.

    Article  PubMed  PubMed Central  Google Scholar 

  31. Candela Marroquín E, Mateos Iglesia N, Palomo Cobos L. Adequacy of medication in patients 65 years or older in teaching health centers in cáceres, spain. [adequacy of medication in patients 65 years or older in teaching health centers in cáceres, spain]. Rev Esp Salud Publica. 2012;86(4):419–34. https://doi.org/10.4321/S1135-57272012000400009.

  32. Gibert P, Cabaret M, Moulis M, Bosson J, Boivin J, Chanoine S, Allenet B, Bedouch P, Gavazzi G. Optimizing medication use in elderly people in primary care: Impact of STOPP criteria on inappropriate prescriptions. Arch Gerontol Geriatr. 2018;75:16–9. https://doi.org/10.1016/j.archger.2017.10.022.

    Article  PubMed  Google Scholar 

  33. Martín Lesende I, Mendibil Crespo I, Maiz López G, Gabilondo Zelaia I, Aretxabaleta Parra JC, Mota GA. Potentiality of STOPP/START criteria used in primary care to effectively change inappropriate prescribing in elderly patients. Eur Geriatr Med. 2013;4(5):293–8. https://doi.org/10.1016/j.eurger.2013.06.006.

    Article  Google Scholar 

  34. Ministerio de Sanidad, Servicios Sociales e Igualdad. Informe anual del sistema nacional de salud 2017. prestación farmacéutica [Web document]. [cited Jun 28, 2019]. <https://www.mscbs.gob.es/estadEstudios/estadisticas/sisInfSanSNS/tablas Estadisticas/InfAnSNS.htm.>.

  35. Cruz-Esteve I, Marsal-Mora JR, Galindo-Ortego G, Galván-Santiago L, Serrano-Godoy M, Ribes-Murillo E, Real-Gatius J. Análisis poblacional de la prescripción potencialmente inadecuada en ancianos según criterios STOPP/START (estudio STARTREC). [potentially inappropriate prescribing in older spanish population according to STOPP/START criteria (STARTREC study)]. Aten Primaria 2017;49(3):166–76. https://doi.org/10.1016/j.aprim.2016.02.013.

  36. Pardo-Cabello AJ, Manzano-Gamero V, Zamora-Pasadas M, Gutiérrez-Cabello F, Esteva-Fernández D, Luna-Del Castillo JdD, Jiménez-Alonso J. Potentially inappropriate prescribing according to STOPP-2 criteria among patients discharged from internal medicine: Prevalence, involved drugs and economic cost. Arch Gerontol Geriatr. 2018;74:150–4. https://doi.org/10.1016/j.archger.2017.10.009.

  37. Santolaya-Perrín R, Jiménez-Díaz G, Galán-Ramos N, Moreno Carvajal MT, Rodríguez-Camacho JM, Sierra-Sánchez JF, Arévalo-Serrano J, Calderón-Hernanz B. A randomised controlled trial on the efficacy of a multidisciplinary health care team on morbidity and mortality of elderly patients attending the emergency department. study design and preliminary results. Farm Hosp. 2016;40(5):371–84. https://doi.org/10.7399/fh.2016.40.5.10465.

  38. Huibers CJA, Sallevelt BTGM, Heij JMJO, O'Mahony D, Rodondi N, Dalleur O, van Marum RJ, Egberts ACG, Wilting I, Knol W, Afd Pharmacoepi & Clinical Pharmacology, Pharmacoepidemiology and Clinical Pharmacology. Hospital physicians' and older patients' agreement with individualised STOPP/START-based medication optimisation recommendations in a clinical trial setting. Eur Geriatr Med 2022;13(3):541–52. https://doi.org/10.1007/s41999-022-00633-5.

  39. Sallevelt BTGM, Huibers CJA, Heij JMJO, Egberts TCG, van Puijenbroek EP, Shen Z, Spruit MR, Jungo KT, Rodondi N, Dalleur O, Spinewine A, Jennings E, O’Mahony D, Wilting I, Knol W. Frequency and acceptance of clinical decision support system-generated STOPP/START signals for hospitalised older patients with polypharmacy and multimorbidity. Drugs Aging 2022;39(1). https://doi.org/10.1007/s40266-021-00904-z.

  40. Gallagher P, Ryan C, Byrne S, Kennedy J, O'Mahony D. STOPP (screening tool of older person’s prescriptions) and START (screening tool to alert doctors to right treatment). consensus validation. Int J Clin Pharmacol Ther 2008;46(2):72–83. https://doi.org/10.5414/CPP46072.

  41. Pérez T, Moriarty F, Wallace E, McDowell R, Redmond P, Fahey T. Prevalence of potentially inappropriate prescribing in older people in primary care and its association with hospital admission: Longitudinal study. BMJ. 2018;14(363):k4524. https://doi.org/10.1136/bmj.k4524.

    Article  Google Scholar 

  42. Damoiseaux-Volman BA, Medlock S, Raven K, Sent D, Romijn JA, van der Velde N, Abu-Hanna A. Potentially inappropriate prescribing in older hospitalized dutch patients according to the STOPP/START criteria v2: A longitudinal study. Eur J Clin Pharmacol. 2021;77(5):777–85. https://doi.org/10.1007/s00228-020-03052-2.

    Article  PubMed  Google Scholar 

  43. Ministerio de Sanidad. Agencia Española de Medicamentos y Productos Sanitarios. Utilización de medicamentos antiulcerosos en españa [Web document]. , 2021 [cited April 28, 2023]. <https://www.aemps.gob.es/medicamentos-de-uso-humano/observatorio-de-uso-de-medicamentos/informes/>.

  44. Mud Castelló F, Mud Castelló S, Rodríguez Moncho MJ, Ivorra Insa MD, Ferrándiz Manglano ML. Detección de prescripciones potencialmente inapropiadas en pacientes ancianos: Estudio descriptivo en dos farmacias comunitarias. Farm Comunitarios. 2014;6(2):20–6. https://doi.org/10.5672/FC.2173-9218.(2014/Vol6).002.04.

    Article  Google Scholar 

  45. Palchik V, Bianchi M, Colautti M, Salamano M, Pires N, Catena JM, Dolza ML, Tassone V, Lillini G, Paciaroni J, Traverso ML. Atención farmacéutica de adultos mayores. aplicación de los criterios STOPP-START. [pharmaceutical care for older adults. application of STOPP-START criteria]. J Healthc Qual Res. 2020;35(1):35–41. https://doi.org/10.1016/j.jhqr.2019.08.003.

  46. Chiam R, Saedon N, Khor HM, A/P. Subramaniam S, Binti Mohmad Nasir SS, Binti Abu Hashim NFI, Tan MP. Potentially inappropriate prescribing in a falls clinic using the STOPP and START criteria. Int J Clin Pharm 2022;44(1):163–71. https://doi.org/10.1007/s11096-021-01329-9.

  47. Ministerio de Sanidad, Servicios Sociales e Igualdad. Informe anual del sistema nacional de salud 2017. demografía y situación de salud [Web document]. [cited Jun 30, 2019]. <https://www.mscbs.gob.es/estadEstudios/estadisticas/sisInfSanSNS/tablasEstadisticas/InfAnSNS.html>.

Download references

Acknowledgements

The authors acknowledge to Ph.D María Jesús Dorta for providing part of the data for this article and to Mr. Jorge Figueroa for his contribution to the English translation of this article.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Author information

Authors and Affiliations

  1. Hospital Universitario Nuestra Señora de La Candelaria. Servicio de Farmacia. Santa Cruz de Tenerife, 38010, Tenerife, Spain

    Silvia González-Munguía

  2. Programa de Doctorado en Ciencias de La Salud. Universidad de La Laguna, 38200, San Cristóbal de La Laguna. Tenerife, Spain

    Silvia González-Munguía

  3. Departamento de Ingeniería Química y Tecnología Farmacéutica. Facultad de Farmacia, Universidad de La Laguna, 38200, San Cristóbal de La Laguna. Tenerife, Spain

    Obdulia Munguía-López & Esther Sánchez Sánchez

Authors
  1. Silvia González-Munguía
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Obdulia Munguía-López
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Esther Sánchez Sánchez
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

SGM conceived and performed the method in RStudio used in the work. She took part in the interpretation of data and participated in the analysis and discussion of the results and drafted the manuscript. OML designed of the work and participated the acquisition and interpretation of data; participated in the discussion of the results, drafted the manuscript and revision of several manuscript versions. ESS helped in the study protocol; participated in the interpretation of data and in the analysis and discussion of the results, drafted the manuscript and revision of several manuscript versions. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Esther Sánchez Sánchez.

Ethics declarations

Ethics approval and consent to participate

This study has been conducted in accordance with the Declaration of Helsinki and has been approved by the Research Ethics and Animal Welfare Committee of the University of La Laguna (CEIBA2017-0255). The need for informed consent was waived by the Research Ethics and Animal Welfare Committee of the University of La Laguna (CEIBA2017-0255).

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

González-Munguía, S., Munguía-López, O. & Sánchez Sánchez, E. Application of STOPP/START criteria in older patients in primary care using RStudio®. BMC Geriatr 24, 782 (2024). https://doi.org/10.1186/s12877-024-05376-5

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12877-024-05376-5

Keywords

BMC Geriatrics

ISSN: 1471-2318

Contact us