Physicians' and pharmacists' perspective on clarity and clinical relevance of absolute contraindications in “Summaries of Product Characteristics”

Abstract

Aims

Previous work has identified several limitations in “Summaries of Product Characteristics” (SmPCs), which are associated with risks for patients. The aim of this study was to evaluate pharmacists' and physicians' interpretation of contraindications in SmPCs and reasons for their nonadherence in clinical routine.

Methods

For 20 commonly missed or ignored absolute contraindications, an anonymous online survey providing 24 clinical example cases (one or two per contraindication) for physicians and pharmacists was developed. Experts in medication safety were asked whether the respective case fulfilled the definition of the contraindication in the SmPC: (a) formally, irrespective of the clinical relevance of the contraindication (17 cases), and (b) whether the contraindication was deemed clinically relevant in each respective case (24 cases).

Results

Twenty‐seven pharmacists and 27 physicians completed the survey. For only one case (1/17; 5.8%) did all experts agree on the same answer option regarding the formal fulfilment of a given contraindication statement. Experts gave heterogeneous answers regarding the interpretation of a contraindication. For instance, among 10 predefined answer options for the contraindication “active liver disease” in the SmPC of simvastatin, every option was selected by at least six experts. In 17/24 (70.8%) clinical example cases a majority of experts agreed on the clinical relevance of a given contraindication. Key reasons for nonadherence to contraindications were “patient monitoring possible”, “lack of alternative treatment” and “acute/severe situations”.

Conclusions

Experts' disagreement on the interpretation of contraindications in SmPCs using clinical example cases indicates that further efforts are needed to improve their usability in clinical routine.

What is already known about this subject

• Contraindications are intended to prevent harmful use of drugs.

• Contraindications are in multiple cases ignored in clinical routine.

• Contraindication statements given in Summaries of Product Characteristics (SmPCs) are sometimes ambiguous and unsuitable for clinical decision making.

What this study adds

• An empiric, survey‐based approach to identify contraindication statements in SmPCs that are difficult to implement in clinical routine.

• Physicians and pharmacists frequently differ in their assessment regarding the applicability and clinical relevance of contraindication statements.

• The study provides user‐based empirical data on why contraindications are frequently ignored in clinical decision making.

1. INTRODUCTION

According to the European Medicines Agency (EMA), contraindications are defined as “situations where a medicinal product must not be given for safety reasons”. ¹ Contraindications can include clinical diagnoses, concomitant diseases, demographics (age, gender), predispositions (metabolism, immunology, genotype), drug–drug combinations and prior adverse reactions to certain drugs. ¹ Failure to adhere to contraindications in clinical practice can lead to serious adverse drug effects resulting in hospitalization or even death. ² , ³

Therefore, descriptions of contraindications, provided in the “Summaries of Product Characteristics” (SmPCs), should be unambiguous, comprehensive and clearly outlined, leaving no room for uncertainty or confusion to prescribing physicians. ¹ However, common clinical experience as well as the literature indicate that in multiple cases this may not be the case as a significant number of SmPCs are ambiguous or poorly described. ⁴ , ⁵ As a consequence, contraindications are sometimes ignored in clinical decision making, which potentially negatively impacts the therapies of patients. ² , ³ , ⁶ However, the reasons for overriding contraindications in clinical practice remain uncertain. Furthermore, even though contraindications are provided in SmPCs, it is up to the treating physicians to find alternatives, potentially leaving patients with inferior treatments. ⁷ For instance, diclofenac is contraindicated for patients suffering from chronic heart failure with New York Heart Association (NYHA) class II–IV in countries of the European Union. ⁸ , ⁹ However, data from clinical practice indicate that, despite the fact that the contraindication was introduced ten years ago in European countries, a significant proportion of patients with heart failure are still taking diclofenac. ⁸ , ⁹ Regardless of the rare clinical situations that may justify the use of contraindicated drugs outside official recommendations, other explanations must be found for the non‐negligible number of contraindicated prescriptions.

Empirical data regarding how physicians and pharmacists interpret and apply contraindication statements in specific clinical cases remain limited. Therefore, the primary objective of this study was to assess physicians' and pharmacists' perspective on the relevance and clarity of contraindications for commonly used drugs as well as reasons for not following contraindications in clinical routine. The insights resulting from this study should add information to improve understanding and implementation of contraindications in a real‐world setting, thereby leading to safer and more effective drug use.

2. METHODS

2.1. Definition of contraindications

For the present study, absolute contraindications refer to section 4.3 of SmPCs. ¹ According to the EMA, contraindications are defined as “situations where the medicinal product must not be given for safety reasons”. ¹

2.2. Selection of contraindications

This analysis is based on a list of the most frequently prescribed drugs in the inpatient care sector, as well as the outpatient care sector in Germany. The generation of the list, as well as the selection process of SmPCs, have been described elsewhere. ¹⁰

Based on the list of these drugs, an expert panel, consisting of clinical pharmacists (W.A., P.D., A.K.S., M.I.ST.) and physicians (L.W., K.F., R.M.) performed a consensus discussion to select specific, clinically highly relevant, but commonly missed or ignored absolute contraindications from German SmPCs. However, tumour drugs are underrepresented on this list, therefore a contraindication statement regarding the use of cisplatin in “renal insufficiency” was added as a well‐established example of clinical relevance for the broad field of oncology. ¹¹ After this selection process, 20 contraindication statements were further used in this study (Supplemental Table S1).

For these contraindications, the drugs' corresponding international SmPCs (UK, USA, Switzerland, EMA) were screened, to exclude that the investigators' observations are mainly attributable to peculiarities of the German language or German SmPCs (Supplemental Table S1). In essence, the drug–disease and drug–drug interactions used in our study are very similar, if not identical, to the contraindication statements in the SmPCs from the EU and other healthcare markets and systems. Phenprocoumon is an exception and served as a clinical example of cases in which international statements for the same drug diverge (Supplemental Table S1).

2.3. Design of the questionnaire

Based on the identified contraindication statements, a structured questionnaire for an anonymous online survey using SoSci Survey (version 3.2.30) ¹² was designed (Supplemental Table S2).

Unless indicated otherwise, the questionnaire was mainly single choice. If contraindications were only deemed relevant by the physicians and pharmacists under certain conditions, the respondents could provide their reasoning regarding the applicability in free‐text fields. The questionnaire was not designed to assess pharmacists' and physicians' knowledge regarding the given contraindications as investigated in other studies, ¹³ but to analyse how they would rate clarity and clinical relevance in specific clinical example cases when consulting the SmPC text. As in clinical routine, the experts were free to do further research, e.g., literature research on the cases, before answering the questions. Essentially, they were asked to answer the questions with consideration and use of all information they deemed relevant.

Based on 20 contraindication statements, clinical example cases were created. For 14 of these statements which require some form of clinical judgement, 17 clinical example cases regarding formal fulfilment of the contraindication were created. Furthermore, 24 example cases were created to assess the pharmacists' and physicians' judgements regarding the clinical relevance of the contraindication statements in different clinical settings.

Depending on the context, the questionnaire included one or two patient cases for each contraindication, asking two main questions:

1. In your opinion, is the definition of contraindication in this example fulfilled? (17 cases) and

2. Is the contraindication clinically relevant [in a specific clinical setting]? (24 cases)

A translated excerpt of the questionnaire is shown in Table 1 (complete questionnaire in Supplemental Table S2).

TABLE 1.

Excerpt from the questionnaire: clarity and clinical relevance of the contraindication diclofenac + heart failure.

Diclofenac + heart failure (NYHA II), one‐time coadministration	Original statement from Summary of Product Characteristics: Diclofenac tablets are contraindicated in “known heart failure (NYHA II–IV)”
Case A	A patient with heart failure (ICD I50.‐heart failure) NYHA II is prescribed a single dose of diclofenac tablets due to acute pain.
	In your opinion, is the definition of contraindication in this example fulfilled?
	A = The definition of the contraindication is formally fulfilled. B = The definition of the contraindication is NOT formally fulfilled.
	Is the prescription clinically relevant contraindicated?
	1 = The contraindication is NOT clinically relevant and does not warrant a warning or evaluation as a medication error. 2 = The contraindication is clinically relevant and warrants a warning. However, the prescription is justifiable under certain conditions. Please explain the condition(s) in the comment field. 3 = The contraindication is clinically relevant, should be evaluated as a medication error, and warrants a warning.
diclofenac + heart failure (NYHA II), continued coadministration	Original statement from Summary of Product Characteristics: Diclofenac tablets are contraindicated on patients with “known heart failure (NYHA II–IV)”
Case B	A patient with heart failure (ICD‐10 I50 Heart failure) NYHA II is prescribed daily administration of diclofenac tablets to treat chronic pain.
	In your opinion, is the definition of contraindication in this example fulfilled?
	A = The definition of the contraindication is formally fulfilled. B = The definition of the contraindication is NOT formally fulfilled.
	Is the prescription clinically relevant contraindicated?
	1 = The contraindication is NOT clinically relevant and does not warrant a warning or evaluation as a medication error. 2 = The contraindication is clinically relevant and warrants a warning. However, the prescription is justifiable under certain conditions. Please explain the condition(s) in the comment field. 3 = The contraindication is clinically relevant, should be evaluated as a medication error, and warrants a warning.

Abbreviations: ICD, International Statistical Classification of Diseases and Related Health Problems; NYHA, New York Heart Association.

For 6 of 17 cases regarding contraindications (i.e., bisoprolol + COPD, bisoprolol + peripheral artery disease [PAD], phenprocoumon + renal insufficiency in the presence of atrial fibrillation, cisplatin + renal insufficiency, simvastatin + active liver disease, valsartan + severe liver insufficiency), more than two answer options were possible regarding formal fulfilment of contraindications (Supplemental Table S2).

The questionnaire was reviewed by each expert on the panel to ensure comprehensive coverage of the most relevant contraindications as well as its own comprehensibility. Five further reviewers (four clinical pharmacists and one clinical pharmacologist) evaluated the questionnaire in a pretest. Thereafter, their suggestions were discussed and afterwards incorporated in the questionnaire.

2.4. Survey study

The experts for the survey were recruited via a mailing list with approximately 140 subscribers from a network of physicians and pharmacists collaborating in the POLAR_MI project, which is addressing medication safety issues in German university hospitals, resulting in a broad spectrum of clinical backgrounds. ¹⁴ The majority of subscribers showed an academic background, having affiliations with universities and university clinics. ¹⁴ The subscribers to the mailing list were invited to pass on the invitation to others.

For the anonymized survey, a corresponding consent waiver of the Ethics Committee of the Friedrich‐Alexander‐Universität Erlangen‐Nürnberg was obtained. Furthermore, the Ethics Committee of the University Hospital Jena took note of the consent waiver.

2.5. Survey analysis

Descriptive statistical analysis was performed to provide a comprehensive summary of collected data. Results include data from all experts and are stratified by pharmacists and physicians to facilitate comparisons between the two professions. Answer patterns in tables were graphically represented in order to allow visualization of the distribution of answers. Majority‐based consensus between experts was defined when more than 50% of experts agreed on the same answer option. The free‐text answers were grouped by an author with experience in qualitative analysis (M.I.SP.) and were all reviewed by a second author (W.A.). The survey was opened July 2021 and closed in February 2022 after reminders were sent in September 2021, December 2021 and January 2022.

2.6. Nomenclature of targets and ligands

Key protein targets and ligands in this article are hyperlinked to corresponding entries in http://www.guidetopharmacology.org, and are permanently archived in the Concise Guide to PHARMACOLOGY 2023/24.

3. RESULTS

The questionnaire was answered by 27 physicians and 27 pharmacists. Of these, 26 (48.1%) were male and 28 (51.9%) were female. Among physicians, 16 (59.3%) were employed at hospitals, 2 (7.4%) were General Practitioners, and 12 (44.4%) were affiliated with universities (multiple answers were possible). Among pharmacists, 15 (55.6%) indicated affiliations with universities, 7 (25.9%) were employed in public pharmacies, and 6 (22.2%) worked at hospitals (multiple answers were possible). The general characteristics of the participants are shown in Table 2.

TABLE 2.

General characteristics of the survey participants.

Gender
Male	26 (48.1%)
Female	28 (51.9%)
Diverse	0 (0.0%)
Age (years)
25–29	20 (37.0%)
30–34	7 (13.0%)
35–39	6 (11.1%)
40–44	3 (5.6%)
45–49	6 (11.1%)
50–54	7 (13.0%)
55–59	3 (5.6%)
60–64	1 (1.9%)
> 65	1 (1.9%)
Occupation
Physician	27 (50.0%)
Specialization (multiple answers possible)
General Practitioner	2 (7.4%)
Hospital	16 (59.3%)
University/research	12 (44.4%)
Other (e.g., pharmaceutical industry)	2 (7.4%)
Pharmacist	27 (50.0%)
Specialization (multiple answers possible)
Public pharmacy	7 (25.9%)
Hospital pharmacy	6 (22.2%)
University/research	15 (55.6%)
Other (e.g., pharmaceutical industry)	4 (14.8%)
Professional experience (years)
< 1	5 (9.3%)
1–5	19 (35.2%)
6–15	14 (25.9%)
16–25	9 (16.7%)
> 26	6 (11.1%)
Not answered	1 (1.9%)

3.1. Expert assessment of whether the definition of the contraindication is formally fulfilled

For only one (i.e., bisoprolol + chronic obstructive pulmonary disease [COPD]) of the 17 cases did all physicians and pharmacists agree on the same answer option whether the patient case formally fulfils the original statement from the SmPC. However, experts did not agree, when there were multiple possible definitions for a contraindicated condition (e.g., COPD Global Initiative for Chronic Obstructive Lung Disease [GOLD] I, II, III or IV), which of the specific definitions is applicable. The answer patterns of physicians and pharmacists are shown in Table 3.

TABLE 3.

Answer patterns of pharmacists, physicians and the total population for the questionnaire, concerning formal fulfilment of the provided contraindications.

Cases with original statement from SmPC	Total population (n = 54)		Pharmacists (n = 27)		Physicians (n = 27)
	Absolute and relative (in %) number of votes for answer options A and B a
	A	B	A	B	A	B
Metamizole + cytostatic treatment for colorectal carcinoma Original statement from SmPC: “disorders of bone marrow function (e.g., after cytostatic treatment)”	35 (66.0)	18 (34.0)	15 (55.6)	12 (44.4)	20 (76.9)	6 (23.1)
Metamizole + iron deficiency anaemia Original statement from SmPC: “diseases of the haematopoietic system”	27 (50.0)	27 (50.0)	11 (40.7)	16 (59.3)	16 (59.3)	11 (40.7)
Diclofenac + heart failure (NYHA II), one‐time coadministration Original statement from SmPC: “known heart failure (NYHA II–IV)”	48 (88.9)	6 (11.1)	25 (92.6)	2 (7.4)	23 (85.2)	4 (14.8)
Diclofenac + heart failure (NYHA II), continued coadministration Original statement from SmPC: “known heart failure (NYHA II–IV)”	52 (96.3)	2 (3.7)	27 (100.0)	0 (0.0)	25 (92.6)	2 (7.4)
Diclofenac + recurrent peptic ulcer, one‐time coadministration Original statement from SmPC: “existing or history of recurrent peptic ulcers or haemorrhage (at least two distinct episodes of proven ulceration or bleeding)”	46 (85.2)	8 (14.8)	25 (92.6)	2 (7.4)	21 (77.8)	6 (22.2)
Diclofenac + recurrent peptic ulcer, continued coadministration Original statement from SmPC: “existing or history of recurrent peptic ulcers or haemorrhage (at least two distinct episodes of proven ulceration or bleeding)”	53 (98.1)	1 (1.9)	27 (100.0)	0 (0.0)	26 (96.3)	1 (3.7)
Amlodipine + decompensated heart failure after myocardial infarction Original statement from SmPC: “haemodynamically unstable heart failure after acute myocardial infarction”	37 (69.8)	16 (30.2)	22 (81.5)	5 (18.5)	15 (57.7)	11 (42.3)
Cotrimoxazole + platelet count under lower limit of normal after chemotherapy Original statement from SmPC: “pathological blood count changes (thrombocytopenia, granulocytopenia, megaloblastic anaemia)”	49 (90.7)	5 (9.3)	25 (92.6)	2 (7.4)	24 (88.9)	3 (11.1)
Metformin + COPD (GOLD stage II) Original statement from SmPC: “acute[ly] or chronic[ally] conditions that may lead to tissue hypoxia, e.g., cardiac or respiratory failure”	32 (60.4)	21 (39.6)	18 (66.7)	9 (33.3)	14 (53.8)	12 (46.2)
Metformin + chronic hypoxemic respiratory failure in the setting of COPD Original statement from SmPC: “acute[ly] or chronic[ally] conditions that may lead to tissue hypoxia, e.g., cardiac or respiratory failure”	49 (92.5)	4 (7.5)	27 (100.0)	0 (0.0)	22 (84.6)	4 (15.4)
Potassium + renal insufficiency, metabolic acidosis and peripheral oedema Original statement from SmPC: “impaired excretory renal function”	50 (92.6)	4 (7.4)	26 (96.3)	1 (3.7)	24 (88.9)	3 (11.1)
Bisoprolol + COPD Original statement from SmPC: “severe chronic obstructive pulmonary disease”	54 (100.0)	0 (0.0)	27 (100.0)	0 (0.0)	27 (100.0)	0 (0.0)
Bisoprolol + PAD Original statement from SmPC: “severe forms of peripheral arterial disease”	52 (98.1)	1 (1.9)	26 (96.3)	1 (3.7)	26 (100.0)	0 (0.0)

Abbreviations: SmPC, Summary of Product Characteristics; NYHA, New York Heart Association; COPD, Chronic Obstructive Pulmonary Disease; GOLD, Global Initiative for Chronic Obstructive Lung Disease; PAD, Peripheral Arterial Disease.

^a Answer options:

A = The definition of the contraindication is formally fulfilled.

B = The definition of the contraindication is NOT formally fulfilled.

Not all participants answered all questions (on average 0.7% unanswered per item).

0.0%–24.9%.

25.0%–49.9%.

50.0%–74.9%.

75.0%–100.0%.

[Correction added on 22 November 2024, after first online publication: The presentation of Table 3 has been corrected in this version.]

For 11 of 17 (64.7%) cases, a majority agreed, whether and when the given contraindications were formally fulfilled, irrespective of their personal assessment of the clinical correctness and relevance of the contraindication (Table 4). In four cases, all 27 pharmacists agreed on the same answer option, whereas all 27 physicians agreed in two cases (Table 3).

TABLE 4.

Consensus of experts on the patient cases.

Possible outcomes	Number of patient cases (n = 24)
	All respondents	Pharmacists	Physicians
Simple majority (>50%) of respondents agreed in their relevance assessment	17 (70.8%)	19 (79.2%)	14 (58.3%)
Type of majority agreement
Clinically relevant a	12 (50.0%)	13 (54.2%)	10 (41.6%)
Clinically relevant under certain conditions	3 (12.5%)	2 (8.3%)	3 (12.5%)
Not clinically relevant	2 (8.3%)	4 (16.7%)	1 (4.1%)
No consensus reached	7 (29.2%)	5 (20.8%)	10 (41.7%)

This table describes for how many of 24 patient cases a consensus could be reached by the experts. A consensus among the questioned experts was defined when a simple majority was given.

It is possible that a simple majority (i.e., 28 of 54 did agree on the same answer option) was not reached among all experts (pharmacists + physicians; n = 54) for a specific patient case. However, subgroups of the whole expert panel reached a simple majority, e.g., 14 of 27 pharmacists agreed on the same answer option while the remaining 13 pharmacists and 27 physicians did not.

^a Clinically relevant: The contraindication is clinically relevant, should be evaluated as a medication error, and warrants a warning.

Clinically relevant under certain conditions: The contraindication is clinically relevant and warrants a warning. However, the prescription is justifiable under certain conditions.

Not clinically relevant: The contraindication is NOT clinically relevant and does not warrant a warning or evaluation as a medication error.

Physicians and pharmacists were asked, under which conditions (e.g., different degrees of severity of a certain disease) they deemed the contraindication to be formally fulfilled, according to the wording provided in the SmPC. Figure 1 shows the variety of answers, for which cases experts deemed the contraindication “active liver disease” according to the SmPC to be fulfilled. Among 10 predefined answer options for the contraindication “active liver disease” in the SmPC of simvastatin, every answer option was selected by at least six experts. Experts gave the most votes to hepatitis, including acute viral hepatitis A, chronic viral hepatitis B and autoimmune hepatitis, to formally fulfil the contraindication “active liver disease”. The experts' interpretation of different contraindication statements from SmPCs are provided in Supplemental Figure S1.

FIGURE 1.

Experts' interpretation of different contraindication statements from SmPCs: simvastatin + active liver disease. Twenty‐seven pharmacists and 27 physicians assessed the clinical applicability of the contraindication statement “active liver disease” in the SmPC of simvastatin. The figure shows the variety of given answers and potential lack of clarity when the contraindication “simvastatin + active liver disease” according to the SmPC is fulfilled. Multiple answers were possible. Abbreviations: AST, aspartate transferase; ALT, alanine transferase; SmPC, Summary of Product Characteristics; ULN, upper limit of normal.

3.2. Clinical relevance

Regarding clinical relevance, there was not a single question which all experts agreed on the same answer option. However, for 17 of 24 (70.8%) clinical example cases, more than 50% of experts agreed on the same answer option (i.e., reached consensus). Pharmacists reached consensus for 19 (79.2%) cases while physicians reached it for 14 (58.3%) cases. Table 5 provides the answer patterns of the participants' ratings regarding the clinical relevance of various contraindications, ranked from highest relevance to lowest relevance.

TABLE 5.

Answer patterns of pharmacists, physicians and the total population for the questionnaire, concerning clinical relevance of the provided contraindications.

Cases with original statement from SmPC	Total population (n = 54)			Pharmacists (n = 27)			Physicians (n = 27)
	Absolute and relative (in %) number of votes for answer options 1, 2 and 3 a
	1	2	3	1	2	3	1	2	3
Diclofenac + heart failure (NYHA II), continued coadministration Original statement from SmPC: “known heart failure (NYHA II–IV)”	1 (1.9)	9 (16.7)	44 (81.5)	0 (0.0)	1 (3.7)	26 (96.3)	1 (3.7)	8 (29.6)	18 (66.7)
Diclofenac + recurrent peptic ulcer, continued coadministration Original statement from SmPC: “existing or history of recurrent peptic ulcers or haemorrhage (at least two distinct episodes of proven ulceration or bleeding)”	0 (0.0)	12 (22.2)	42 (77.8)	0 (0.0)	5 (18.5)	22 (81.5)	0 (0.0)	7 (25.9)	20 (74.1)
Ciclosporin + dabigatran Original statement from SmPC: “in combination with dabigatran”	3 (5.6)	9 (16.7)	42 (77.8)	2 (7.4)	5 (18.5)	20 (74.1)	1 (3.7)	4 (14.8)	22 (81.5)
Simvastatin + active liver disease Original statement from SmPC: “active liver disease”	3 (5.9)	15 (29.4)	33 (64.7)	2 (8.0)	8 (32.0)	15 (60.0)	1 (3.8)	7 (26.9)	18 (69.2)
Bisoprolol + PAD Original statement from SmPC: “severe forms of peripheral arterial disease”	8 (15.1)	5 (9.4)	40 (75.5)	3 (11.1)	3 (11.1)	21 (77.8)	5 (19.2)	2 (7.7)	19 (73.1)
Valsartan + severe liver insufficiency Original statement from SmPC: “severe liver insufficiency”	6 (12.0)	10 (20.0)	34 (68.0)	3 (12.0)	5 (20.0)	17 (68.0)	3 (12.0)	5 (20.0)	17 (68.0)
Cisplatin + renal insufficiency Original statement from SmPC: “pre‐existing renal insufficiency”	1 (2.0)	22 (43.1)	28 (54.9)	0 (0.0)	11 (42.3)	15 (57.7)	1 (4.0)	11 (44.0)	13 (52.0)
Bisoprolol + COPD Original statement from SmPC: “severe chronic obstructive pulmonary disease”	10 (18.5)	6 (11.1)	38 (70.4)	8 (29.6)	1 (3.7)	18 (66.7)	2 (7.4)	5 (18.5)	20 (74.1)
Metformin + chronic hypoxemic respiratory failure in the setting of COPD Original statement from SmPC: “acute or chronic conditions that may lead to tissue hypoxia, e.g., cardiac or respiratory failure”	8 (15.1)	10 (18.9)	35 (66.0)	3 (11.1)	3 (11.1)	21 (77.8)	5 (19.2)	7 (26.9)	14 (53.8)
Zolpidem + sleep apnoea syndrome Original statement from SmPC: “sleep apnoea syndrome”	9 (16.7)	14 (25.9)	31 (57.4)	5 (18.5)	4 (14.8)	18 (66.7)	4 (14.8)	10 (37.0)	13 (48.1)
Phenprocoumon + renal insufficiency in the presence of atrial fibrillation Original statement from SmPC: “conditions where the risk of bleeding outweighs the potential therapeutic benefit, e.g., […] manifested renal insufficiency”	8 (15.4)	16 (30.8)	28 (53.8)	3 (11.5)	7 (26.9)	16 (61.5)	5 (19.2)	9 (34.6)	12 (46.2)
Amlodipine + decompensated heart failure after myocardial infarction Original statement from SmPC: “haemodynamically unstable heart failure after acute myocardial infarction”	10 (18.9)	14 (26.4)	29 (54.7)	4 (14.8)	6 (22.2)	17 (63.0)	6 (23.1)	8 (30.8)	12 (46.2)
Potassium + renal insufficiency, metabolic acidosis and peripheral oedema Original statement from SmPC: “impaired excretory renal function”	6 (11.3)	23 (43.4)	24 (45.3)	1 (3.7)	10 (37.0)	16 (59.3)	5 (19.2)	13 (50.0)	8 (30.8)
Citalopram + amitriptyline Original statement from SmPC: “with concomitant use of drugs known to cause prolongation of the QT interval”	0 (0.0)	36 (66.7)	18 (33.3)	0 (0.0)	20 (74.1)	7 (25.9)	0 (0.0)	16 (59.3)	11 (40.7)
Levofloxacin + epilepsy Original statement from SmPC: “epilepsy”	5 (9.3)	27 (50.0)	22 (40.7)	2 (7.4)	13 (48.1)	12 (44.4)	3 (11.1)	14 (51.9)	10 (37.0)
Hydrochlorothiazide + gout (without gout drug treatment) Original statement from SmPC: “gout”	11 (20.4)	18 (33.3)	25 (46.3)	7 (25.9)	7 (25.9)	13 (48.1)	4 (14.8)	11 (40.7)	12 (44.4)
Hydrochlorothiazide + gout (with gout drug treatment) Original statement from SmPC: “gout”	11 (20.4)	23 (42.6)	20 (37.0)	5 (18.5)	11 (40.7)	11 (40.7)	6 (22.2)	12 (44.4)	9 (33.3)
Cotrimoxazole + platelet count under lower limit of normal after chemotherapy Original statement from SmPC: “pathological blood count changes (thrombocytopenia, granulocytopenia, megaloblastic anaemia)”	10 (18.9)	26 (49.1)	17 (32.1)	6 (22.2)	13 (48.1)	8 (29.6)	4 (15.4)	13 (50.0)	9 (34.6)
Metamizole + cytostatic treatment for colorectal carcinoma Original statement from SmPC: “disorders of bone marrow function (e.g., after cytostatic treatment)”	12 (22.2)	31 (57.4)	11 (20.4)	4 (14.8)	19 (70.4)	4 (14.8)	8 (29.6)	12 (44.4)	7 (25.9)
Hydrochlorothiazide + severe renal insufficiency (GFR < 30 mL/min) Original statement from SmPC: “severe renal impairment”	16 (29.6)	29 (53.7)	9 (16.7)	10 (37.0)	13 (48.1)	4 (14.8)	6 (22.2)	16 (59.3)	5 (18.5)
Diclofenac + recurrent peptic ulcer, one‐time coadministration Original statement from SmPC: “existing or history of recurrent peptic ulcers or haemorrhage (at least two distinct episodes of proven ulceration or bleeding)”	26 (48.1)	12 (22.2)	16 (29.6)	14 (51.9)	3 (11.1)	10 (37.0)	12 (44.4)	9 (33.3)	6 (22.2)
Metformin + COPD (GOLD stage II) Original statement from SmPC: “acute or chronic conditions that may lead to tissue hypoxia, e.g., cardiac or respiratory failure”	26 (49.1)	14 (26.4)	13 (24.5)	14 (51.9)	6 (22.2)	7 (25.9)	12 (46.2)	8 (30.8)	6 (23.1)
Diclofenac + heart failure (NYHA II), one‐time coadministration Original statement from SmPC: “known heart failure (NYHA II–IV)”	30 (55.6)	15 (27.8)	9 (16.7)	18 (66.7)	5 (18.5)	4 (14.8)	12 (44.4)	10 (37.0)	5 (18.5)
Metamizole + iron deficiency anaemia Original statement from SmPC: “diseases of the haematopoietic system”	37 (68.5)	16 (29.6)	1 (1.9)	19 (70.4)	7 (25.9)	1 (3.7)	18 (66.7)	9 (33.3)	0 (0.0)

Abbreviations: SmPC, Summary of Product Characteristics; NYHA, New York Heart Association; PAD, peripheral arterial disease; COPD, chronic obstructive pulmonary disease; GFR, glomerular filtration rate; GOLD, Global Initiative for Chronic Obstructive Lung Disease.

^a Answer options:

1 = The contraindication is NOT clinically relevant and does not warrant a warning or evaluation as a medication error.

2 = The contraindication is clinically relevant and warrants a warning. However, the prescription is justifiable under certain conditions. Please explain the condition(s) in the comment field.

3 = The contraindication is clinically relevant, should be evaluated as a medication error, and warrants a warning.

Not all participants answered all questions (on average 1.4% unanswered per item).

0.0%–24.9%.

25.0%–49.9%.

50.0%–74.9%.

75.0%–100.0%.

With respect to duration of treatment, both professions made a clear distinction with respect to clinical relevance and applicability of the contraindication statements between single dose (i.e., one‐time coadministration) and continued coadministration. The contraindication diclofenac + heart failure (NYHA II) was rated among the least important, when only a one‐time administration of diclofenac is given to a heart failure patient, whereas it was ranked most important in continued therapy (compare Tables 3 and 5).

In many clinical example cases the experts deemed the contraindication statements to be formally applicable, but not clinically relevant. For instance, 48 (88.9%) of the experts agreed that the one‐time coadministration of diclofenac for a patient with heart failure (NYHA II) formally fulfils the contraindication statement from the respective SmPC (which does not differentiate one‐time coadministration and chronic coadministration), but only 9 (16.7%) of the experts deemed the contraindication in case of one‐time coadministration to be clinically relevant, i.e. considered the one‐time coadministration a medication error warranting a warning (compare Tables 3 and 5).

Overall, physicians and pharmacists provided similar ratings when asked for clinical relevance. However, pharmacists tended to rate the clinical relevance of the evaluated contraindications higher than physicians (Table 5).

For 6 of 17 (35.3%) contraindications, physicians and pharmacists were also asked, under which conditions (e.g., different degrees of severity of a certain disease) they deemed the contraindication to be clinically relevant, according to the wording provided in the SmPC. The respective ratings are provided in Supplemental Figure S2.

3.3. Free‐text answers

Both professional groups rated “patient monitoring possible” as the most important condition, under which a prescription is considered clinically acceptable despite a formal (i.e., technical) contraindication (Supplemental Figure S3).

Additionally, for the contraindications involving an antibiotic drug (i.e., levofloxacin + epilepsy, co‐trimoxazole + platelet count under lower limit of normal after chemotherapy), “lack of alternative treatment” and “acute/severe situations” were named as factors justifying a prescription despite a formally absolute contraindication. Table 6 summarizes the main reasons for tolerating a clinically relevant contraindication.

TABLE 6.

Main reasons for “tolerating” a clinically relevant contraindication.

Reason for tolerating a contraindication	Total number of votes (multiple answers possible)
Patient monitoring possible	188
Lack of alternative treatment	55
Acute/severe situations	39
Drug monitoring possible	22
[Short] duration of treatment	21
Other	99

4. DISCUSSION

To serve their purpose, SmPCs should list contraindications in a clear and comprehensive manner, so that in a given clinical setting all healthcare professionals come to the same (correct) clinical decision regardless of their specific clinical background. Our survey based on common real‐world examples shows that this may often not be the case.

4.1. General findings

In this anonymous survey of physicians and pharmacists regarding their interpretation of the clinical applicability of contraindication statements, only for one contraindication did all experts agree on the same answer option. For broad or imprecise clinical terms used in the statements, such as “liver disease” or “impaired excretory renal function”, the respondents differed in the clinical thresholds they deemed to be applicable. Furthermore, in many clinical example cases, the healthcare professionals deemed the contraindication statements to be formally applicable, but not clinically relevant.

From these observations, two main conclusions can be drawn: (1) Many contraindications are worded ambiguously in the SmPCs, leaving too much room for interpretation. This results in different opinions and treatment decisions by experts independently assessing the same patient case, as other studies were able to show in the past. ⁵ , ¹⁵ (2) The perceived clinical relevance of different contraindications in clinical routine settings varies substantially.

4.2. Main reasons for tolerating a clinically relevant contraindication

The analysis of the detailed answers and free text of the individual clinical example cases provided some insights regarding the underlying mechanisms. “Patient monitoring” (e.g., ECG monitoring), “lack of alternative treatment”, “drug monitoring” or “short duration of treatment” could be valid reasons for tolerating absolute contraindications listed in SmPCs. With respect to duration of treatment, both professions made a clear distinction with respect to clinical relevance and applicability of the contraindication statements between single dose (i.e., one‐time coadministration) and continued coadministration. The contraindication diclofenac + heart failure (NYHA II) was rated among the least important when only a one‐time administration of diclofenac is given to a heart failure patient, whereas it was ranked most important in continued therapy.

Following this result, we suggest considering a more detailed description of certain contraindications in SmPCs to improve acceptance and implementation, for example by differentiating between one‐time use and continuous therapy.

4.3. Differentiation of absolute and relative contraindication

A clearer differentiation between absolute and relative contraindications in SmPCs could also enhance the acceptance and implementation rates of contraindication alerts. The EMA and US Food and Drug Administration (FDA) might take varied approaches to this issue. Specifically, the absolute number of drug contraindications is lower in the US market than in German SmPCs, which we already noted in previous works based on analyses with 158 and 228 SmPCs, respectively (Supplemental Table S1). ⁵ , ¹⁵ This could be due to a clinically more sensitive differentiation between “truly” absolute and relative contraindications in US SmPCs. Simple definitions such as “liver disease” or “impaired renal function” may additionally be suitable to raise alertness and ensure that any possible uncritical use is discouraged, but it may also result in overalerting and undertreatment.

In line with this, the ambiguity of contraindication statements like “liver disease” may also explain some of the variation in the reported prevalence of medication errors. ¹⁵

4.4. Exclusion criteria in randomized controlled trials

Another reason for nonadherence to contraindication statements such as “active liver disease” in SmPCs might result from the EMA's definition of contraindications, on which this study is based. According to the EMA, patient populations not studied in the randomized controlled trials should not be mentioned in the contraindication section of SmPCs. ¹ Therefore, lack of data alone is not a reason for a contraindication statement. However, “if […] patients have been excluded from studies due to a contraindication on grounds of safety, they should be mentioned in this section [i.e., section 4.3 contraindications].” ¹ This could be another reason why respective contraindications are ignored in clinical practice. For instance, several types of liver disease may appear as a contraindication in SmPCs, resulting from restrictive exclusion criteria in randomized controlled trials, such as excluding patients with mild elevations of transaminases. Clarifying this origin in the contraindication statements of the SmPC could facilitate their interpretation.

4.5. Different interpretation of statements by pharmacists and physicians

Our findings could lead to an explanation for the different acceptance rates of pharmacists' recommendations to physicians, as not all recommendations are always followed. ¹⁶ , ¹⁷

Comparing the interpretation of contraindications by pharmacists and physicians, a consensus could be reached by pharmacists in 19 (79.2%) cases, compared to physicians who reached consensus in 14 (58.3%) cases. These differences in interpretation of contraindication statements may be attributable to aspects of training and the respective roles of physicians and pharmacists in the prescription process.

4.6. Limitations

Our sample size in participants and examples might not have been large enough to identify all possible types of issues and to provide generalizable statements about the proportion of possible problematic contraindication statements. However, empiric knowledge from other interview studies shows that an effective sample size (information saturation) can be expected with the sample size of 54 participants (27 in each sub group) in this study. ¹⁸

5. CONCLUSION AND FUTURE OUTLOOK

This online survey among physicians and pharmacists provides clear empirical data indicating that the description of contraindications in SmPCs may frequently leave too much room for interpretation. Furthermore, the clinical relevance of a contraindication depends highly on the clinical situation (e.g., long‐term vs. single application), which may be insufficiently accounted for in some current SmPCs.

Our study offers a case‐based empirical approach as an instrument for the evaluation of the clinical relevance and applicability of contraindications in clinical settings for different healthcare professions and where the statements lack precision. Future use cases could potentially be the easy modification and application of the survey tool in other countries or including and comparing larger participant samples of specialized professions. Such surveys might also help to identify unanticipated sources of error.

To improve the clinical usability of contraindication statements, and SmPCs in general, physicians and pharmacists should more consistently communicate cases where statements in SmPCs lack clarity or are perceived as clinically misleading to the manufacturers or drug authorities. This could be easily organized as a feature of the reporting systems and platforms that are already in use for the reporting of unexpected side effects. ¹⁹ , ²⁰

AUTHOR CONTRIBUTIONS

W.A., M.I.SP. and R.M. wrote the manuscript. W.A., P.D., A.K.S., L.W., K.F., M.I.ST., R.M. and M.F.F. designed the research. W.A., P.D., A.K.S., L.W., K.F., M.I.ST., R.M. and M.F.F. performed the research. W.A., M.I.SP. and L.J.‐P. analysed the data. All authors revised the manuscript for important content.

CONFLICT OF INTEREST STATEMENT

W.A., M.I.SP., L.J.‐P., A.K.S., L.W., K.F. and R.M. declare no conflict of interest. M.I.ST. has received consultancy fees from Heumann Pharma and Pharbil Pharma. P.D. has received lecture fees and reimbursement of travel costs from AstraZeneca. M.F.F. has received consultancy fees from Boehringer Ingelheim and lecture fees from Janssen‐Cilag. He has received third‐party funds for research projects at his institution by Boehringer Ingelheim and Heidelberg Pharma Research GmbH. M.F.F. received an earmarked financial contribution for the first award of the MSD Germany Health Award 2021.

Supporting information

Figure S1. Experts' interpretation of different contraindication statements from SmPCs. The figure shows the variety of given answers and potential lack of clarity when the contraindications “bisoprolol + COPD”, “bisoprolol + PAD”, “phenprocoumon + renal insufficiency in the presence of atrial fibrillation”, “cisplatin + renal insufficiency” and “valsartan + severe liver insufficiency” according to the SmPC are fulfilled. Results for the contraindication “simvastatin + active liver disease” are shown in Figure 1. Abbreviations: COPD, chronic obstructive pulmonary disease; PAD, peripheral artery disease; SmPC, Summary of Product Characteristics.

Figure S2. Expert assessment of the severity or stage at which they deemed the contraindication to be clinically relevant. The figure demonstrates the experts' ratings as of which disease stages the given contraindications were deemed as clinically relevant. Abbreviations: COPD, chronic obstructive pulmonary disease; PAD, peripheral artery disease.

Figure S3. Free text comments from physicians and pharmacists on conditions under which the experts deemed a contraindication not to be clinically relevant (and did not consider the administration of the drug to be a medication error but justifiable). This figure demonstrates the experts' ratings, under which conditions the given contraindication is justifiable.

Table S1. Comparison of German SmPCs with SmPCs/Prescribing Information from UK, USA, Switzerland and the EMA. The 17 contraindication statements of German SmPCs evaluated in this study were matched with the corresponding SmPCs/Prescribing Information from the UK, USA, Switzerland and the EMA. The comparison of contraindication statements is based on the following scale: 0 = not present; 1 = contraindication; 2 = warning. An evaluation “12/14”, as indicated for the UK, means that 14 of 17 German contraindication statements could be matched with corresponding UK SmPCs and 12 of these 14 statements constitute absolute contraindications.

Table S2. Questionnaire. The whole questionnaire of the study.

Andrikyan W, Sponfeldner MI, Jung‐Poppe L, et al. Physicians' and pharmacists' perspective on clarity and clinical relevance of absolute contraindications in “Summaries of Product Characteristics”. Br J Clin Pharmacol. 2025;91(3):829‐840. doi: 10.1111/bcp.16331

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request and in compliance with German data protection laws.