Implementation of electronic distribution of Direct Healthcare Professional Communications by the Danish Medicines Agency: a survey study of physicians’ experiences and preferences

Abstract

Background:

The efficient distribution of Direct Healthcare Professional Communications (DHPCs) is crucial for ensuring healthcare providers promptly receive important new safety information.

Objectives:

This study aimed to evaluate the implementation of the electronic distribution of DHPCs by the Danish Medicines Agency (DKMA) and to assess how future safety communication can be improved.

Design:

We conducted a web-based cross-sectional survey among Danish physicians using a self-administered questionnaire.

Methods:

DKMA sends DHPCs to healthcare professionals via an electronic mailbox called e-Boks which is linked to the unique personal identifier. To evaluate DKMA’s distribution of DHPCs, participants were asked about awareness, frequency and barriers of reading, preferred distribution channel and overall satisfaction. To further identify potential improvements, respondents were asked about general preferences regarding sender and channels of safety information in addition to which information sources they use to keep up-to-date.

Results:

A total of 2238 physicians completed the survey corresponding to a response rate of 26% based on the total target population. The total awareness was 81%. Compared to previous research, awareness of GPs increased from 66% to 82%, and the percentage of GPs who rarely or never read DHPCs decreased from 33% to 20%. In addition, our study revealed a preference for receiving DHPCs electronically through e-Boks as opposed to workplace-delivered postal letters, and a preference for the DKMA over pharmaceutical companies as the sender of DHPCs. One-third of the respondents were either ‘dissatisfied’ or ‘very dissatisfied’ with the current solution. A professional mailbox and point-of-care alerts when prescribing may complement the primary distribution channel to strengthen the uptake. Additionally, existing information sources already frequented by the target group may be used to communicate safety information.

Conclusion:

The DKMA’s electronic distribution of DHPCs suggests an improvement and may serve as inspiration for other agencies. However, the considerable dissatisfaction calls for further improvements.

Plain language summary

Evaluating physicians’ preferences and experiences with email delivered safety information of medicines

Why was this study done? A direct healthcare professional communication (DHPC) serves the purpose of providing timely important safety information about medicine to physicians. This is crucial for the safe and effective use of medicines. The main goal of our study was to see how well the Danish Medicines Agency (DKMA) is doing with sending DHPCs via an electronic mailbox called e-Boks, and to find ways to make this communication even better in the future. e-Boks is linked to your Danish unique personal identifier, which remains with you even if you move or change your email address. What did the researchers do? We surveyed a group of Danish physicians using an online questionnaire. The questions covered topics like how often they read DHPCs, any difficulties they face, their preferred way of receiving them, and their overall satisfaction with the current system. What did the researchers find? Of the 2238 physicians who responded, 81% were aware of DHPCs, marking a significant increase compared to previous years. Most preferred receiving DHPCs electronically through e-Boks rather than through postal letters at their workplace. They also favored getting this information from the DKMA instead of pharmaceutical companies. However, a third of the respondents were not happy with the current system. What do the findings mean? The findings suggest that the DKMA’s method of electronically distributing DHPCs is well-received by physicians and has improved their awareness and reading of these important safety communications. To further enhance the effectiveness of safety communication, it is recommended to diversify the electronic channels used. This could include professional emails, pop-up messages when they prescribe medicine, and making use of existing information platforms frequently visited by physicians. The study indicates that the DKMA’s approach could serve as a model for other agencies, but also highlights the need for ongoing efforts to refine the system.

Introduction

A medicine can only be authorised if its benefits outweigh the risks. Ideally, the benefits of a medicinal product should exceed the risks by the greatest achievable margin.^1,2 When a new risk has been identified, appropriate regulatory action is taken to minimise the risk and improve the risk–benefit balance of the medicinal product. ³

In the EU, risk minimisation measures (RMMs) include routine measures (e.g. the product information, the pack size and whether or not the medicinal product is subject to medical prescription) and additional measures. Routine RMMs are mandatory and apply to every authorised medicine. In exceptional cases, for selected important risks, routine RMMs are insufficient to manage the risk and/or improve the risk–benefit balance of a product, and thus additional risk minimisation measures (aRMM) are required. ¹ The most common aRMMs are educational materials (including prescriber brochures and patient alert cards) and Direct Healthcare Professional Communications (DHPCs). ⁴

A DHPC is a communication intervention to inform healthcare professionals (HCPs) of the need to take certain actions (e.g. monitoring instructions for preventing adverse reactions) or adapt their practices in relation to a medicinal product (e.g. restriction of an indication or new contraindications). ⁵ In contrast to the product information where information is provided on a website for HCPs to retrieve, the DHPC is delivered directly to HCPs.

In May 2019, the Danish Medicines Agency (DKMA) replaced the marketing authorisation holder (MAH) as the sender of DHPCs. This change was first announced as part of a growth plan for the Danish life science industry issued by the Danish government. A primary reason for introducing electronic distribution was poor awareness and reading of DHPCs but also to reduce the burden on the pharmaceutical industry. Previously, MAHs invested significant time and effort in retrieving the specific names and addresses of national HCPs, often through consultancies. This step has been eliminated since the DKMA took over the responsibility. As a result of this transition, the annual fee that all MAHs pay to the DKMA for covering general authority tasks was slightly increased in 2019 to DKK 70/year (approximately 9 EUR/year). Additionally, the DKMA received 0.75 full-time equivalent resources to manage the distribution of DHPCs, which is estimated to cover the current expenses.

Following the transition, the DHPC letter remained the same including its formal features and industry signatories. However, instead of postal mail, DHPCs began to be distributed electronically, directly to HCPs’ personal electronic mailboxes, known as e-Boks. e-Boks is a personal electronic mailbox mandatory for all Danish citizens aged 15 and above, enabling secure communication between public authorities, citizens and businesses. Unlike ordinary emails, e-Boks is linked to your Danish unique personal identifier, ensuring that it remains with you even if you move or change your email address. Similar to ordinary mailboxes that can be used for both personal and professional communication, e-Boks can also be linked to CVR numbers (Chamber of Commerce numbers) for professional or business purposes.

The identification of relevant HCPs for a specific DHPC is based on data from the Danish registry of HCPs with an authorisation to work as an HCP. Similarly, the DKMA can address private medical clinics, private hospitals and pharmacies through their registered CVR numbers (Chamber of Commerce number) in e-Boks owned by the clinics (professional e-Boks). Of note, GPs in Denmark either have their medical practice or share a practice with other GPs. Furthermore, the Danish Regions (representing Danish hospitals) receive the DHPC by email, which they distribute to relevant departments at the hospitals. Finally, a link to the DHPC is located under the respective products in a publicly accessible digital database of available medicines, owned by the Danish Association of the Pharmaceutical Industry (Lif).

One of the reasons for changing the distribution of DHPCs was a study from 2013 referred to in the DKMA’s annual pharmacovigilance report, showing that only two out of three GPs were aware of DHPCs (N = 422), but of these, one in three read them rarely or never. ⁶

The main change in the distribution of the DHPC after the DKMA took over the responsibility in May 2019, was the change of communication channel (electronic vs postal letter) and change of the sender (MAH vs National Competent Authority (NCA)).

Regarding communication channels, a 2011 French study found that postal letters were the preferred method (42%), with email/Internet following at 25%. ⁷ Subsequent research in the Netherlands, however, indicates a preference among HCPs for electronic means of distribution, such as email.^8–10 Electronic distribution is noted for its user-friendliness, offering direct links to detailed information on safety issues and the ability to highlight the specific safety concern and associated drug in the message header. ⁹

Previous research by de Vries et al. showed that the most preferred senders of safety information were NCAs and professional bodies. ¹⁰ They concluded that to improve safety communication strategies, it should be clear to the receiver that the information comes from the NCA or another preferred sender, such as a professional body. The study was conducted among nine European NCAs including Denmark but was limited by low sample size as only 25 GPs participated from Denmark.

In another study by de Vries et al., ¹¹ which investigated the sources through which HCPs become aware of specific safety issues, they recommend, in alignment with European Medicines Agency (EMA) guidelines, ⁵ that NCAs should consider using various information sources as this is supposed to strengthen the uptake of the information. This approach is further supported by Møllebæk and Kaae who recommend that emergent risk communication should involve further integration into existing structures for clinical recommendations and risk management. ¹² A study among 1580 Danish GPs about information-seeking behaviour revealed that medical websites were used weekly by 91% and drug information websites by 74%. However, no details about which specific websites they preferred were provided, and the study was restricted to GPs only. ¹³

To our knowledge, DKMA is the first NCA to use electronic distribution of DHPCs. Hence, there is a need to evaluate how this solution works in real life. The study aimed to describe the implementation of electronic distribution of DHPCs by the DKMA in terms of awareness, reading, barriers to reading and satisfaction, and further to describe general preferences on safety communication including sender, communication channel and information sources. Based on this information, we assessed whether the communication of DHPCs has improved since the DKMA took over the responsibility of distributing DHPCs, and how future safety communication about medicines can be improved with particular consideration of DHPCs.

Methods

Study design and target population

We conducted a web-based cross-sectional open survey among Danish GPs, oncologists, haematologists and gastroenterologists/hepatologists using a self-administered questionnaire (see Supplemental Material). Oncologists, haematologists and gastroenterologists/hepatologists were selected because they received the most DHPCs in Denmark in 2022. In addition, GPs were selected because they are the largest group of physicians in Denmark who in 2019 totalled 3402 full-time professionals. ¹²

The survey was conducted between January and March 2023.

The reporting of this study conforms to the CHERRIES statement. ¹⁴

Recruitment

We used e-Boks to invite all Danish GPs, oncologists, haematologists and gastroenterologists/hepatologists to participate in the survey. All physicians were identified from the Danish registry of HCPs (https://autregweb.sst.dk/authorizationsearch.aspx). A reminder was sent 1 week after the first invitation. The survey was also announced on the website of the Danish Society of Clinical Oncology, the website of the Danish Society of Gastroenterologists and Hepatologists. In addition, the Danish Organization of General Practitioners, which is a union of GPs, announced the survey on their website and Facebook as well. The survey was not announced on the website of the Danish Society of Haematology because the website is only used for professional purposes and the traffic on the website is generally low.

Participation was anonymous and voluntary. According to the Danish National Center for Ethics, surveys and interviews not involving human biological material do not require ethical approval or participant consent (https://nationaltcenterforetik.dk/ansoegerguide/overblik/hvad-skal-jeg-anmelde).

No incentives were offered for participation.

Development of the questionnaire

The survey questions were inspired by the research Strengthening Collaboration for Operating Pharmacovigilance in Europe (SCOPE) Work Package 6, especially regarding preferences of sender and channel. The SCOPE Work Package project was sponsored by the European Commission and focused on risk communications. ¹⁰

Since poor awareness and reading of DHPCs were among the reasons for changing their distribution and were therefore crucial for the evaluation, these questions were phrased to facilitate comparison with previous research. ⁶ Additionally, since the changes in distribution were related to the alteration of the sender and the channel, questions about these aspects were included. From stakeholder meetings prior to the implementation of the new distribution method for DHPCs, we learned that there was some resistance to using the personal e-Boks. Consequently, we included questions about satisfaction and potential barriers.

The questionnaire was developed and reviewed by a broad team of experts including researchers in pharmacoepidemiology, regulatory science and communication, and staff working with DHPCs at DKMA.

Cognitive pre-testing was performed in three physicians according to the ‘Thinking aloud method’. ¹⁵ Following the pre-test and review, minor changes were made to reduce the number of questions and to increase the clarity and understanding of each question. Pre-testing indicated that it would take approximately 10 min to complete the questionnaire.

LimeSurvey version 2.67.2 (Limesurvey GmbH, Hamburg, Germany, http://www.limesurvey.org, an open-source survey tool) was used for the web-based survey.

Outcome assessment

Communication model

The Communication Sequence model (Figure 1) conceptualises communication as a totality of five constitutive parts in a temporal sequence: sender, message, medium, recipient and effect. ¹⁶

Figure 1.

Communication sequence model.

This model reflects a passive receiver where information is pushed through by a sender, also called push information. ¹⁷ It is a model that describes how DHPCs are communicated where information is pushed by a competent authority to an HCP. In our study, we focused on the two features of communication, sender and media/channel as these features capture the main changes in the distribution of DHPCs. ‘Senders’ refer to persons/organisations that distribute the information (e.g. NCAs, pharmaceutical companies, researchers, press), and ‘channels’ refer to the medium/mode of dissemination through which information is disseminated (e.g. personalised letter, email, social media).

HCPs are not only passive receivers of information, they are also active information-seeking. To capture this aspect of communication, we use the term ‘information source’, which is defined as a person, thing or place that is accessed by an HCP to keep up-to-date with safety information on medicines. This type of information is also called pull information. ¹⁷ Data about information sources were collected to explore the use of other information sources to complement the DHPC and to give input to a broader discussion about how future safety communication about medicines can be improved.

Outcomes

The extent of improvement in communication of DHPCs, since the DKMA took over the responsibility of distributing DHPCs through e-Boks, was assessed. Respondents were asked about, general awareness of DHPCs, overall satisfaction with receiving DHPCs through e-Boks, whether they read the DHPCs they receive, and barriers to reading DHPCs (refer to the Supplemental Material for an overview of how study objectives and survey questions are linked). In addition, they were asked about their preferred channel (electronic vs postal letter).

To measure overall satisfaction, we used a 5-point Likert scale ranging from ‘very dissatisfied’ to ‘very satisfied’ complemented by a free text field. We used a 5-point ordinal scale to assess the frequency of reading ranging from ‘none/few’ to ‘most/all’. For determining barriers to reading and the preferred method of receiving DHPCs, we opted for closed ended (fixed alternative) multiple-response choices (refer to the Supplemental Material for details about the response options).

To further elucidate how future safety communication about medicines can be improved in Denmark, respondents were asked about general preferences regarding sender and channels of safety information on medicines. Moreover, they were asked which information sources they use to keep up-to-date with safety information on medicines.

To measure preferences for four predefined senders and nine predefined channels of safety information about medicine, we employed a 5-point Likert scale ranging from ‘very negative’ to ‘very positive’. While we used a 4-point ordinal scale to determine the frequency with which they use various predefined sources of information about medicines ranging from ‘daily’ to ‘never’. All questions about general preferences of safety information included a free text field.

Except for the question about barriers to reading, every question included a ‘do not know’ option.

Data analysis

We performed descriptive analyses on all outcomes, evaluating each survey question individually by frequency and percentage. Free text answers were collected in an Excel sheet and then categorised inductively. Initially, one researcher read through all responses and organised them into several main categories (e.g. recipient, layout, channel) and subcategories. Subsequently, a second researcher reviewed the categorisation performed by the first researcher. Finally, any disagreements were discussed between the two researchers to reach a consensus. All analyses were also conducted excluding pensioners and for each specialty separately (see Supplemental Material).

All data handling and analysis were performed using SPSS (IBM Corp, released 2017, IBM SPSS Statistics for Windows, Version 25.0, Armonk, NY, USA).

Data were reported in an aggregate format only, without any personal identifiers.

Missing data were not imputed.

Results

Characteristics of respondents

In total, 2238 physicians completed the whole survey corresponding to an overall response rate of 26% (2238/8738). For individual questions, the response rate may be higher (Figure 2). It is worth noticing that the response rate is based on the total target population of the four specialties that was selected for this survey.

Figure 2.

Flowchart of respondents.

*We approached the full target population.

**The question about barriers to reading was optional.

Of the 2238 physicians who completed the survey, 76% (1694/2238) were GPs, 5% (105/2238) oncologists, 3% (59/2238) haematologists and 3% (76/2238) gastroenterologists/hepatologists. The remaining 14% (304/2238) had a specialty that was unable to categorise into the four specialties of the target population. The majority in this group indicated they were pensioners using the ‘Other’ response option, but they did not specify their specialty within the four predefined response options.

About a quarter (28%, 635/2238) were pensioners and not practising physicians. If not otherwise specified, pensioners are included in the results presented below (see Supplemental Material for results excluding pensioners).

Although there were slightly more elderly respondents (34% vs 30% for the age group 70+) compared to the target population (Table 1), respondents were considered largely representative of the target population regarding gender and age.

Table 1.

Characteristics of participating physicians indicating gender and age.

Gender and Age	Respondents Background (Q1–Q4) N = 2551*		Respondents** DHPC (Q5–Q11) N = 1930***		Respondents**** general preferences (Q12–Q14) N = 2238*****		Target population N = 8738
	n	%	N	%	n	%	n	%
Gender
Female	1133	44	911	47	1022	46	4107	47
Male	1414	55	1016	53	1214	54	4631	53
Other	4	<1	3	<1	2	<1	–	–
Age (years)
30–39	88	3	70	4	83	4	437	5
40–49	494	19	438	23	457	20	2184	25
50–59	447	18	383	20	410	18	1748	20
60–69	571	22	446	23	502	22	1748	20
70+	931	37	580	30	772	34	2621	30

^*Respondents with unknown age = 20.

^**Only those who were aware of DHPCs answered these questions.

^***Respondents with unknown age = 13.

^****Also respondents that were unaware of DHPCs answered these questions.

^*****Respondents with unknown age = 14.

DHPC, Direct Healthcare Professional Communication.

Implementation of electronic distribution of DHPCs through the personal e-Boks

Awareness

Awareness of DHPCs was a condition for answering questions about DHPCs. The total awareness including all respondents (including respondents with unknown specialty) was 81% (1959/2415) of which 90% (1754/1959) did receive four or more DHPCs the past year. Awareness was lowest among GPs (82%, 1487/1818) and highest among haematologists (99%, 64/65).

Satisfaction

Slightly more respondents were dissatisfied than satisfied with the current solution where DHPCs are sent to the personal e-Boks. A total of 34% (642/1932) of the respondents were ‘dissatisfied’ or ‘very dissatisfied’, 32% were neutral and 32% (635/1932) were satisfied or very satisfied (Figure 3).

Figure 3.

Satisfaction with receiving DHPCs through the personal e-Boks: How satisfied are you with receiving safety information on medicines through the e-Boks (on a 5-point Likert scale from very dissatisfied to very satisfied)? N = 1932.

DHPC, Direct Healthcare Professional Communication.

We also included an optional question where respondents in a free text field could express their dissatisfaction in relation to receiving DHPCs in their personal e-Boks. The comments were categorised into three major themes:

• Channel – personal e-Boks is inappropriate for professional communication (N = 396). Respondents find it inappropriate and disturbing to receive professional information in the personal e-Boks, and many prefer to receive the information by mail in a professional mailbox.

• Recipient – irrelevant information (N = 201). Respondents reported that they receive DHPCs for medicine they never prescribe.

• Layout – difficult to navigate (N = 95). Respondents find it difficult to navigate the DHPC and suggested that this could be improved by a short summary at the beginning of the DHPC and by providing more information in the subject field of the mail (e.g. name of the medicinal product).

Reading and barriers to reading

In total, 20% (391/1932) of the respondents reported that they read none or few of the DHPCs they received, whereas 37% read most/all (Figure 4). The most frequent reason, as reported by 55% (1059 out of 1932) of respondents, was that they do not prescribe the medication in question. This reason was reported for 69% (269/391) of those who read none or a few of the DHPCs they received. Other common barriers to reading a DHPC were its perceived irrelevance to daily practice (20%, 385/1932) and the length of the DHPC (13%, 242/1932). Barriers related specifically to the e-Boks included that it is too difficult/time-consuming to open a DHPC via e-Boks (11%, 216/1932) and that many do not read work-related information received through the personal e-Boks (11%, 214/1932).

Figure 4.

Reading: How many of the DHPCs you receive do you read (on a 5-point scale from 0% to 20% (none/few) to 81% to 100% (most/all))? N = 1932.

DHPC, Direct Healthcare Professional Communication.

Of note, 31 respondents reported that they do not read the DHPC because DHPCs are perceived as marketing/advertising material.

Preferred channel – postal letter or email

The current solution, where GPs receive the DHPC in both the personal e-Boks and the professional e-Boks, was preferred by 8% of the GPs. The personal e-Boks as a single channel, preferred by 33% and the professional e-Boks alone, preferred by 39%, were almost equally popular (Figure 5(a)).

Figure 5.

(a) General practitioners’ preferences towards communication channels for DHPCs: Do you prefer to receive DHPCs as a postal letter sent to your workplace or sent as an email to your e-Boks? N = 1197. (b) Oncologists’, haematologists’ and gastroenterologists’/hepatologists’ preferences towards communication channels for DHPCs: Do you prefer to receive DHPCs as a postal letter sent to your workplace or sent as an email to your personal e-Boks? N = 240.

DHPC, Direct Healthcare Professional Communication.

Currently, for the vast majority of GPs, DHPCs are sent to both the personal e-Boks and the professional e-Boks. For the vast majority of the three other specialties, the DHPCs are only sent to the personal e-Boks. Hence, the response options for the question about preferences towards communication channel for DHPCs are different for GPs (Figure 5(a)) and the other three specialties (Figure 5(b)).

General preferences of safety communication

General preferences of safety communication covered three features of communication: sender, media/channel and information source (in case of information-pull). Familiarity with DHPCs was not a condition for answering questions about general preferences of safety communication.

The highest valued sender of safety information was the DKMA and independent researchers with a median at ‘Positive’ (Table 2), whereas the least valued sender was pharmaceutical companies with a median at ‘Neutral’.

Table 2.

Frequency distribution.

Senders	Very negative (= 1)		Negative (= 2)		Neutral (= 3)		Positive (= 4)		Very positive (= 5)		Median	Interquartile range (IQR)
	n	%	n	%	N	%	n	%	n	%
Danish Medicines Agency (n = 2149)	33	1.5	63	2.9	539	25.1	876	40.8	638	29.7	Positive	Neutral – Very positive
Independent researchers (n = 1884)	49	2.6	116	6.2	771	40.9	630	33.4	318	16.9	Positive	Neutral – Positive
The Danish Medicines Agency in collaboration with pharmaceutical companies (n = 2006)	73	3.6	195	9.7	849	42.3	665	33.2	224	11.2	Neutral	Negative – Neutral
Pharmaceutical companies (n = 2069)	173	8.4	516	24.9	946	45.7	368	17.8	66	3.2	Neutral	Negative – Neutral

Preferences towards various senders of safety communication: How do you value the following senders of safety information about medicines (on a Likert scale from 1 (very negative) to 5 (very positive))?

The highest rated channel of safety communication was the professional e-Boks with a median at ‘Positive’ (Table 3). Other possible alternative channels of safety communication (median = neutral) were Point-of-care alerts (e.g. pop-up notification at the time of prescribing or dispensing), Email (professional), Daily meeting in the section and Personal e-Boks.

Table 3.

Frequency distribution.

Channels	Very negative (= 1)		Negative (= 2)		Neutral (= 3)		Positive (= 4)		Very Positive (= 5)		Median	Interquartile range (IQR)
	n	%	n	%	n	%	n	%	n	%
Professional e-Boks (n = 1374)	89	6.5	144	10.5	391	28.5	512	37.3	238	17.3	Positive	Neutral – Positive
Point-of-care alerts (n = 1854)	256	13.8	265	14.3	456	24.6	632	34.1	245	13.2	Neutral	Negative – Positive
Professional email (n = 1701)	196	11.5	294	17.3	476	28.0	531	31.2	204	12.0	Neutral	Negative – Positive
Personal e-Boks (n = 2184)	423	19.4	347	15.9	420	19.2	661	30.3	333	15.2	Neutral	Negative – Positive
Daily meeting in the section (n = 915)	120	13.1	149	16.3	315	34.4	237	25.9	94	10.3	Neutral	Negative – Positive
Mobile app (n = 1659)	499	30.1	469	28.3	481	29.0	174	10.5	36	2.2	Negative	Very negative – Neutral
TV and radio (n = 1930)	593	30.7	573	29.7	626	32.4	127	6.6	11	0.6	Negative	Very negative – Neutral
Mobile phone text, e.g. SMS (n = 1946)	984	50.6	622	32.0	199	10.2	109	5.6	32	1.6	Very negative	Very negative – Negative
Social media, e.g. Twitter, LinkedIn (n = 1710)	868	50.8	533	31.2	239	14.0	61	3.6	9	0.6	Very negative	Very negative – Negative

Preferences towards various channels of safety communication: How do you value the following channels for safety information about medicines (on a Likert scale from 1 (very negative) to 5 (very positive))?

By far the most preferred information source on a daily/weekly basis for keeping up-to-date with safety information was Medicin.dk/Pro.medicin.dk (Table 4), which is a publicly accessible digital database with information on available medicines that is owned by the Danish Association of the Pharmaceutical Industry (Lif) and edited by an independent board of editors. The least used information sources were company representatives, The Danish Medicines Council and The Regional Medicines Committees which were never used by 63%, 50% and 40% of the respondents, respectively.

Table 4.

Information sources: How often do you use the following options to keep your medicines knowledge up-to-date?.

Information sources	Daily (%)	Weekly (%)	Monthly or less (%)	Never (%)	Do not know (%)
Medicin.dk/Pro.medicin.dk (a website providing information about medicine to HCPs; N = 2247)	48	23	22	4	3
Summary of product characteristics (N = 2241)	2	14	58	21	5
Rational Pharmacotherapy (a monthly newsletter from the Danish Health Authority; N = 2242)	1	15	66	13	4
The Regional Medicines Committees (the committee develops regional guidelines for the primary sector; N = 2239)	<1	5	40	40	14
The Danish Medicines Council (provide guidance about new medicines for use in the hospital sector; N = 2238)	<1	3	30	50	17
International and national clinical guidelines from medical associations (N = 2243)	9	28	45	11	6
The Danish Medicines Agency’s (DKMA) website/newsletter (N = 2240)	1	9	46	33	10
The Journal of the Danish Medical Association (N = 2248)	1	46	42	9	3
Company representatives (N = 2238)	<1	2	28	63	7

HCP, healthcare professional.

Discussion

This study is the first of its kind to investigate the electronic distribution of DHPCs by an NCA.

Based on the results on awareness, reading and barriers to reading DHPCs, in addition to preferences towards sender and communication channel, we are now able to answer the question: has the communication of DHPCs improved since the DKMA took over the responsibility of distributing DHPCs? Moreover, the results on satisfaction yield information on the need for further improvement. Lastly, the results on general preferences of safety communication provide information about how further safety communication may be improved.

Has the communication of DHPCs improved after the implementation of electronic distribution?

Awareness, reading and barriers to reading

In our study, 82% of GPs were aware of DHPCs, and 20% read none or few of the DHPCs they received. This is slightly lower than the 96% awareness that was reported in a previous study conducted in Denmark ¹⁰ ; however, only 25 Danish GPs participated in that study. Our results are in line with a study conducted in the Netherlands where 72% of 233 GPs reported awareness of DHPCs. ⁹ Compared to a study from 2013 referred to in the DKMA’s annual pharmacovigilance report, ⁶ awareness among GPs increased from 66% to 82%, and the percentage of GPs who rarely or never read DHPCs decreased from 33% to 20%. This suggests an improvement in physicians’ awareness and reading of DHPC since DKMA took over the responsibility for their distribution.

Notably, the most common reason for not reading a DHPC was that the physician never prescribed the medicine in question. This reason was reported for 69% of those who read none or a few of the DHPCs they received. While this is a valid reason, it highlights the need for better targeting of the intended audience before distribution. However, it should be noted that even if the physician does not prescribe the medication, they may still encounter patients who experience side effects from it, making the DHPC relevant to their practice. Nonetheless, for reading to be a meaningful outcome measure, the DHPC must contain information that is relevant to the physician. The challenge is to reach those for whom the DHPC is relevant and avoid those for which it is not. Regulators may tend to focus on reaching all HCPs at the expense of low specificity, that is, many receive the DHPC who perhaps should not have received it. This problem is not limited to Denmark. In a study conducted in the Netherlands, the majority (58%) of the HCPs indicated that they read only the DHPCs that contained information that was relevant to them. ⁹

Sender of DHPCs

One of the findings from the 2013 DKMA pharmacovigilance report was that pharmaceutical companies acting as senders pose a barrier to the implementation of DHPCs because it is difficult for physicians to distinguish between objective information about medicine and other commercial material they receive from these companies. ⁶ The change of sender was supported by several previous studies from the EU and the United States which generally show that independent sources like NCAs and professional bodies are preferred over pharmaceutical companies.^9,10,12,18 Our results are consistent with these findings as the DKMA was the highest-valued sender of safety information, and the pharmaceutical companies were ranked lowest. While this result was expected in light of previous research, it confirms DKMA’s status as a trusted source, which is crucial for the efficacy of the communication.^12,19

Although the DKMA replaced the MAH as a sender, the pharmaceutical companies still sign the DHPCs, and for a few respondents (1.6%, 31/1932) the DHPC is still perceived as marketing/advertising material.

In conclusion, the change of the sender from the pharmaceutical industry to the DKMA is an improvement as few respondents perceive the DHPC as marketing/advertising material, and the perceived association of DHPCs with pharmaceutical companies has been reduced.

Preferred channel to receive DHPCs

Consistent with previous research,^8–10 our findings indicate a distinct preference for the electronic dissemination of DHPCs as opposed to traditional postal mail. For specialties other than GPs, about two-thirds preferred the current channel (email to your personal e-Boks) over the prior approach of postal letters delivered to their workplace. Among GPs, four out of five expressed a preference for receiving DHPCs through electronic means, whether through their personal e-Boks, their professional e-Boks, or a combination of both, rather than via postal mail.

However, when presented with a choice between receiving emails in a professional e-Boks versus a personal e-Boks, a marginally higher number of respondents favour the professional e-Boks option. Today, GPs receive the DHPCs in their personal e-Boks and their professional e-Boks. Noteworthy, the professional e-Boks could replace the personal e-Boks. Therefore, the response options included both the current solution, the previous solution and each of the two e-Boks on their own. When GPs were asked about which communication channel they prefer, only 8% selected the current solution, and almost an equal number chose the personal e-Boks or the professional e-Boks despite the pronounced dissatisfaction with the personal e-Boks. Among those respondents who were ‘dissatisfied’ or ‘very dissatisfied’, only 7% (31/470) preferred the personal e-Boks compared to 64% (222/349) among those who were ‘satisfied’ or ‘very satisfied’ indicating that the personal e-Boks is a cause of dissatisfaction. This suggests that the best strategy to improve satisfaction does not necessarily involve choosing one option over the other, but rather, a solution that offers both alternatives.

To summarise, the communication of DHPCs has improved owing to:

• A preference for electronic delivery of DHPCs via personal e-Boks as opposed to workplace-delivered postal letters,

• A preference for the DKMA over pharmaceutical companies as the sender of DHPCs, and finally,

• An increase in awareness and reading of DHPCs since the responsibility of sending DHPCs electronically via e-Boks was undertaken by the DKMA.

How can future safety communication about medicines be improved?

Despite the communication of DHPCs has improved substantially, one-third of respondents still report being ‘dissatisfied’ or ‘very dissatisfied’. This raises a critical question: Is the current practice where the DKMA sends DHPCs electronically to personal mailboxes good enough? Is it adequate?

Given the differences in physicians’ preferences, some level of dissatisfaction with DHPCs is likely unavoidable. Defining an ‘acceptable’ level of dissatisfaction and establishing a benchmark for success in this context is challenging. Moreover, from a regulatory perspective, a certain degree of dissatisfaction might be tolerable, provided that the DHPCs are read and understood by their intended audience. However, the considerable level of dissatisfaction with the current system suggests that an evaluation of the DHPC distribution process is warranted to identify potential areas for improvement.

Our results show that a major challenge with the current setup is related to dissatisfaction with receiving professional information in the personal e-Boks, which is also one of the main barriers to the reading of DHPCs. Many participants expressed that they would prefer to receive DHPCs in their ordinary professional mailbox, and not through e-Boks. A professional email is also ranked high among other possible alternative channels of safety communication. However, it is difficult to ensure that you reach the complete target population using ordinary email as a channel for DHPCs as these are not mandatory and people may change their email addresses. Hence, from a regulatory perspective, ordinary emails should not be used as the primary communication channel of DHPCs but may be used as a supplement to strengthen the uptake, as shown by previous research. ⁸ An advantage of the e-Boks is complete access to all Danish physicians, and since e-Boks is linked to the Danish unique personal identifier, it stays with you, even if you move or change your email address. As stated previously, a solution that offers both alternatives, personal or professional e-Boks, may be an improvement for more than one-third of the GPs who prefer professional e-Boks.

Another way to enhance safety communication is to ensure that it reaches only those individuals for whom the DHPC is relevant, as irrelevance often leads to dissatisfaction. One of the reasons why many respondents perceive the information in DHPCs as irrelevant is because they do not prescribe the medication that the DHPC pertains to. To ensure that the information in DHPCs is useful and adapted to the target audience, it is advised to involve HCPs’ organisations in the preparation of DHPCs. ⁵ This approach is also beneficial in addressing concerns about the excessive length of DHPCs, which has been identified as a major barrier to their reading. However, adherence to this recommendation at the national level is lacking, and to our knowledge, it is not followed by the EMA.

At present, it is not possible to disseminate information to a more specific level than the specialty. Considering the differences in the need for safety information due to varying prescribing practices among physicians within the same specialty (e.g. oncology), it is difficult to reach only the individuals for whom the information is relevant. An alternative and more feasible approach to reduce the frustration in relation to receiving irrelevant information would be to make it easier to determine the relevance of the information. As suggested by some respondents (N = 20), this may be improved by providing more information in the subject field of the mail (e.g. brand name and active substance).

Rather than focusing on the respondents’ dissatisfaction with the current setup, we can instead examine their preferences. By examining preferences, we can improve safety communication by complementing, but not necessarily replace the e-Boks. According to the guideline on good pharmacovigilance practices on safety communication, a DHPC is a communication intervention by which important safety information is delivered directly to individual HCPs. ⁵ Regulatory authorities should adhere to the guidelines, which may make it challenging to accommodate respondent preferences.

The results on general preferences towards communication channels reveal that ‘daily meeting in the section’, emails (professional) and point-of-care alerts (e.g. pop-up notification at the time of prescribing or dispensing) are possible channels to complement the e-Boks. Consistent with a 2012 study on Dutch HCPs, Danish HCPs do not prefer to receive safety information through text messaging and social media. ⁹

At present, the DKMA complements the e-Boks to disseminate DHPCs by sending emails to the Danish Regions (representing Danish hospitals) which then distribute these to relevant departments at the hospitals. We also know that there are examples of DHPCs being presented at morning conferences at Danish hospitals. However, it remains unknown to which extent morning conferences are used to communicate safety information from regulatory authorities, as well as whether we reach more physicians by sending emails to the Danish Regions.

A less intrusive way to reach the target population is to utilise the information sources that HCPs already use. When considering alternative information sources to disseminate safety information, it is important to bear in mind that the DHPC is often disseminated in situations when there is a need to take immediate action in relation to a medicinal product, for instance in case of suspension of a marketing authorisation for safety reasons. Our research findings indicate that the DKMA’s website/newsletter is not a viable option due to the infrequent usage reported by the respondents. This limitation also applies to the use of clinical guidelines, which are updated annually or even less frequently. Besides publishing the DHPC on the DKMA’s website, it is also published at Pro.medicin.dk, which is a publicly accessible digital database of available medicines owned by the Danish Association of the Pharmaceutical Industry (Lif). Our study confirms that Pro.medicin.dk is a suitable information source as it is the most frequently used information source to keep physicians’ medicines knowledge up-to-date. It should be noted though that it is unknown whether they read the DHPC when they visit the website. Our study also reveals that the most obvious alternative information source to further reach the target population is Rational Pharmacotherapy as this is published by another Danish authority, the Danish Health Authority.

Implications for regulators

Regulatory authorities are required to adhere to guidelines that define DHPCs as interventions for delivering crucial safety information directly to individual HCPs. A significant challenge in this process is ensuring that DHPCs reach all HCPs for whom the information is relevant to their clinical practice. This task often involves a trade-off: aiming for complete reach can result in low specificity, leading to the distribution of DHPCs to many HCPs for whom the information is irrelevant. This not only leads to frustration at an individual level but may also cause information fatigue which undermines the effectiveness of regulatory communication. Therefore, it is imperative for regulators to balance the need for thorough dissemination with the responsibility of minimising the spread of extraneous information, thereby safeguarding the integrity and efficacy of healthcare communication.

Based on our findings, we have several recommendations aimed at improving safety communication in general, and in particular to other NCAs who consider to take over the responsibility from the MAH to distribute DHPCs electronically:

• Reduce the perceived association of DHPC with pharmaceutical companies.^9,10,12,18

• Collaborate with HCPs’ organisations to ensure the relevancy and usefulness of the information delivered to the target audience. ⁵

• Make it easy to navigate in the information to facilitate the assessment of its relevance; for example, by incorporating an informative heading.

• Consider integrating professional email as an additional communication channel for disseminating DHPCs.^8,10

• If the target audience is HCPs working at hospitals, consider engaging with the management of relevant hospital departments to communicate the safety information at the regular meetings in the department.

• If feasible, explore the potential use of pop-up notifications during prescribing or dispensing processes. ¹⁰

• Investigate the feasibility of utilising existing information sources already accessed by the target group for disseminating safety communication.^5,11,12

Strengths and limitations

Our study has several strengths. It is the largest study so far on this topic. In a previous study conducted in nine European countries in 2015 (i.e. DK, ES, HR, IE, IT, NL, NO, SE and the UK), the included number of GPs represented between 0.3% (for UK) and 3.7% (for IE) of the total number of GPs in the various countries. Our study is based on a significantly larger sample size of 1694 GPs representing 22% (1694/7735) of the total population of GPs in Denmark, and is representative in terms of gender and age.^9,10 Moreover, a team with different areas of expertise took part in the development of the questionnaire, which was subsequently pre-tested to establish face validity.

While the study provides meaningful insights, it is necessary to interpret the results with several limitations in mind. Firstly, it remains unknown whether the change in the distribution of the DHPC improved patient safety. RMMs, such as the DHPC, are primarily implemented to reduce risks. It is clear that the goal of such measures is to lessen the occurrence or intensity of adverse reactions.

Secondly, representativeness of respondents with respect to other characteristics than gender, age and specialty is not known. It is important to note that we used the same methodology for this survey as for distributing DHPCs. Those who accepted the invitation may also be more likely to read DHPCs and more satisfied with the current distribution. This may have biased results in a more positive direction.

Thirdly, while acknowledging that other HCPs, such as nurses and pharmacists, also contribute to risk minimisation, a limitation of our study is that it exclusively includes physicians. Moreover, the inclusion of only four specialties may limit the extrapolation to other physicians.

Conclusion

Transitioning from the distribution of DHPCs via postal letters by MAHs to electronic distribution by the DKMA was associated with increased awareness and reading of DHPCs. Physicians prefer electronic distribution to workplace-delivered postal letters, and also the DKMA over pharmaceutical companies as the sender. Hence, it is concluded that the DKMA’s electronic distribution of DHPCs suggests an improvement. To our knowledge, the DKMA stands out as the first NCA to adopt electronic distribution of DHPCs and may serve as inspiration for other agencies.

However, many respondents report dissatisfaction due to perceived irrelevance and navigational difficulties in DHPCs. To minimise these issues, we propose collaborating with HCP organisations to ensure relevancy and incorporating informative headings for easier navigation.

To strengthen the uptake of safety communication in general, we recommend using various electronic communication channels, such as professional emails and point-of-care alerts, and supplementing these with existing information sources already frequented by the target group.