Podcast-Based Medical Education and Content Alignment with National Education Bodies

Abstract

Introduction

Despite the rapid expansion of medical education podcasts in the past decade, there are few efforts to characterize the landscape of available content for specific medical specialties. We trialed a method of rigorous characterization for the field of neurology.

Materials and Methods

Using a censoring date of July 25, 2022, we queried the top three podcast platforms for neurology education podcasts: Apple Podcasts, Spotify, and Google Podcasts. We characterized podcasts based on total number of episodes, episode release frequency, target audience, and affiliation type. We characterized individual episodes by length and primary content area as defined by the Accreditation Council for Graduate Medical Education (ACGME) and American Board of Psychiatry and Neurology (ABPN). We compared content availability with content recommendations from these education bodies.

Results

We identified 30 podcasts sharing 1772 episodes and totaling 46,287 min. The ACGME subspecialties most frequently covered were vascular neurology (5082 total min, 11%), neuroimmunology (4,406, 10%), and neuromuscular diseases (3,771, 8%). Subjects that were underrepresented included palliative neurology (89 min, 0.2%), neuropathology (95 min, 0.2%), and bioethics (171 min, 0.4%). The coverage of ABPN examination topics varied substantially from the content distribution for the examination.

Discussion

The current landscape of neurology education podcasts features heterogeneous coverage of topics and varies considerably from recommended distribution of content by national education bodies. As podcasts have tremendous potential in supplementing neurology education, characterizing available content may help various stakeholders in the neurology education pipeline optimize the use of this e-learning modality.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40670-023-01971-0.

Introduction

Asynchronous, on-demand educational resources have become increasingly commonplace in medical education [1]. Among different e-learning modalities, podcasts have emerged as a format boasting accessibility and portability; learner-centric features including ability to pause, repeat, and alter playback speeds [2]; allowance for multitasking [3], for example, while driving [4]; and international reach [5]. The creators of clinically oriented podcasts are diverse, including medical students [6], residents [7], medical societies [8], national academies [9], and even pharmaceutical companies [5]. Some podcasts are content-oriented, while others focus on diagnostic reasoning (e.g. the popular “The Clinical Problem Solvers” podcast [10]) or on recent journal articles [11]. Learner-reported satisfaction and self-reported learning outcomes are generally favorable [3, 12], while studies with objective measurement of learning outcomes are very limited but suggest at least non-inferiority compared to other modalities [3]. With the growing popularity of education podcasts, important guidance has emerged on many stages in their development, from initial conception to technical details to marketing [13, 14].

Among a growing number of open educational resources (OER) [15], podcasts hold particular potential in addressing the global digital divide in health education. Not only can podcasts avail clinical content to countries without ready access to medical education, the format also benefits from great heterogeneity among developers, including those outside of traditional medical education institutions. Given abundant data on the importance of diversity and representation among educators in health education [16], the diversity of podcast developers allows for varied viewpoints on the same topics. This advantage may be especially important for non-clinical subjects such as health equity or ethics.

The format of podcasts aligns well with learning theories related to self-directed learning. However, unique among OERs is the finding that podcast audiences report a connection to a community of listeners [17]. Indeed, due to the more conversational nature of many podcasts as compared to traditional pre-recorded lectures, the pedagogical advantages of podcasts may rely on specific learning theories less relevant for other OERs. One example is the situated learning theory for podcasts that use an immersive clinical reasoning format [18]. Regarding the design of podcast episodes themselves, experts have advised alignment with the well-established Cognitive Theory of Multimedia Learning and adult learning theories [19].

The development and dissemination of education podcasts is unregulated [20], as any individual can create and upload them with minimal equipment and technical knowledge [6, 7]. This has allowed for the rapid expansion of podcasts of various styles that may appeal to different learners with unique preferences, but the lack of regulatory oversight is concerning. There have been a few efforts to characterize the landscape of podcast content for specific disciplines, including emergency medicine [21, 22] and surgical subspecialties [23], which provide a bird’s eyes view of available content.

The characterization of available content has two important implications. First, it provides information on asymmetric topic coverage, which allows podcast developers to prioritize creating future content that fill specific gaps [24]. Second, it provides a snapshot to educators of e-learning available to their learners. This allows them to curate podcast content to add to existing curricula or bolster other parts of learning environments to address underrepresented topics. Despite the value of such information, there are few efforts to characterize content for individual specialties, and none comparing such content with the recommendations of relevant education bodies. Here, we demonstrate one method for rigorous education content characterization for the field of neurology, which may be applied to other specialties.

Materials and Methods

Using a final censoring date of July 25, 2022, we queried the top three podcast platforms for neurology clinical education podcasts: Spotify (Stockholm, Sweden 2006), Apple Podcasts (Cupertino, CA 2012), and Google Podcasts (Mountain View, CA 2018). We used the search terms “neurology,” “neurology education,” and “neuroscience education” to capture podcasts targeting an audience aiming to learn general neurology content as opposed to specific subspecialty content. From a list of 150 podcasts identified as the top 50 podcasts for each platform as ranked by their respective algorithms, we applied exclusion criteria to select relevant neurology education podcasts for review (Fig. 1). Specifically, duplicates were removed. Podcasts that were not well-established (arbitrary selected as < 5 episodes) or those featuring predominantly very short episodes less likely to be effective in conveying knowledge (arbitrarily selected as average < 3 min) were excluded. Finally, podcasts with a target audience other than clinicians (as deemed by the first and senior author based on podcast summary and episode descriptions), for example, patients or research scientists, were excluded.

Fig. 1.

Flow diagram for podcast inclusion and exclusion

We characterized each podcast by total number of episodes, episode release frequency, affiliation, and affiliation type (academic society, private company, academic journal, medical center or hospital, or none). For each podcast, we characterized all episodes by length and content type based on episode title, the latter in two ways. The first used categories as defined by IV.B.1.(c).(2) of Accreditation Council for Graduate Medical Education (ACGME) Program Requirements for Graduate Medical Education in Neurology [25] that included neurologic subspecialties, neuropalliative care, and bioethics. The second used categories as defined by the Dimension 1 Neurologic Disorders and Topics of the 2021 American Board of Psychiatry and Neurology (ABPN) Continuing Certification/Maintenance of Certification Examination in Neurology [26]. Percentages represented by each topic were calculated using a denominator of the total number of minutes that could be classified. We compared these percentages to those listed in the ABPN Dimension 1 list along with variance. Notably, we excluded “neuroimmunologic and paraneoplastic CNS disorders,” as the title of podcast episodes did not allow for differentiation of this category from the overlapping “demyelinating diseases” and “neuro-oncologic disorders” categories. Otherwise, for consistency, we established a set of classification rules a priori to maximize objectivity among authors (Appendix 1).

Results

Our query identified 30 podcasts sharing 1772 episodes totaling 46,287 min (Appendix 2). Podcast developers included representatives of academies and societies (n = 9), medical institutions (n = 5), private organizations (n = 5), academic journals (n = 3), and unspecified entities (n = 6). The number of total podcast episodes per year increased steadily, from 13 in 2008 to 300 in 2021 (Fig. 2). In terms of ACGME categories, a total 7,463 min were unclassified, while the greatest number of minutes were in vascular neurology (5,082, 11%), neuroimmunology (4,406, 10%), and neuromuscular medicine (3,771, 8%) (Fig. 3). In terms of ABPN examination categories (Table 1), the variance between the percentage of podcast minutes dedicated to a topic compared to examination coverage ranged from − 4.1% (headache and pain disorders) and − 2.8% (metabolic diseases, nutritional deficiency states, and disorders due to toxins, drugs, and physical agents) to 4.1% (vascular neurology) and 3.5% (neuro-ophthalmologic and neuro-otologic disorders).

Fig. 2.

Number of neurology podcast episodes by categories as defined by Accreditation Council for Graduate Medical Education (ACGME) Program Requirements for Graduate Medical Education in Neurology, by year

Fig. 3.

Distribution of neurology education podcast minutes by categories as defined by Accreditation Council for Graduate Medical Education (ACGME) Program Requirements for Graduate Medical Education in Neurology, with a censoring date of July 25, 2022

Table 1.

Total neurology education podcast minutes as categorized by the American Board of Psychiatry and Neurology (ABPN) 2021 Neurologic Disorders and Topics Blueprint, as compared with content distribution by the ABPN

Neurologic disorders and topics	% Exam coverage (Avg.)	% Podcast minutes (absolute number of minutes)	Variance
01. Headache and pain disorders	8–12% (10%)	5.9% (2105)	− 4.1%
02. Epilepsy and episodic disorders	8–12% (10%)	8.1% (2862)	− 1.9%
03. Sleep disorders	3–5% (4%)	1.8% (628)	− 2.2%
04. Genetic and developmental disorders	6–8% (7%)	6.2% (2198)	− 0.8%
05. Vascular neurology	8–12% (10%)	14.1% (5002)	4.1%
06. Neuromuscular diseases	8–12 (10%)	10.3% (3656)	0.3%
07. Movement disorders	8–12% (10%)	10.3% (3650)	0.3%
08. Neuroimmunologic and paraneoplastic disorders of the CNS	8–12% (10%)	12.4% (4406)	2.4%
09. Neuroinfectious diseases	2–4% (3%)	3.4% (1215)	0.4%
10. Brain and spinal trauma and spinal diseases	2–4% (3%)	2.1% (751)	− 0.9%
11. Neuro-ophthalmologic and neuro-otologic disorders	2–4% (3%)	6.5% (2318)	3.5%
12. Metabolic diseases, nutritional deficiency states, and disorders due to toxins, drugs, and physical agents	3–5% (4%)	1.2% (412)	− 2.8%
13. Neuro-oncologic disorders	3–5% (4%)	2.3% (812)	− 1.7%
14. Behavioral neurology and neurocognitive disorders	7–9% (8%)	8.1% (2866)	0.1%
15. Psychiatric disorders	1–2% (1.5%)	7.4% (2627)	1.4%
16. Autonomic nervous system disorders	1–2% (1.5%)	0.1% (30)	− 1.4%
17. Normal structure, process, and development through the life cycle	1–2% (1.5%)	0.3% (135)	− 1.2%

Discussion

Our study found great heterogeneity among content coverage for neurology topics as defined by the ACGME Guidelines IV.B.1.c).(2).(b) and (c) [25] ranging from 0.2 to 11% of total content by time. In particular, there is little coverage on topics in neuropalliative care and bioethics even though these subjects may be particularly well-suited to the audio format [27]. We hypothesize that neuropalliative care and bioethics may be neglected due to either a lack of expertise among podcast developers or a lack of awareness regarding their importance. The latter has been increasingly recognized as an issue in medical training [28].

Our study also found substantial variance between the distribution of podcast content and that recommended by the ABPN Continuing Certification/Maintenance of Certification Examination in Neurology 2021 examination. The subspecialties that were most underrepresented in podcasts compared with recommended board examination coverage were sleep neurology and headache and pain disorders. We suspect that this reflects these specialties’ underrepresentation in neurology training in general—neither is mandated as a required rotation across neurology residency programs [29] and the workforces for these specialties may also be declining [30, 31]. One could argue that specialties featured less prominently in clinical training should be asymmetrically featured in supplemental e-learning. The most underrepresented topic not associated with a neurologic subspecialty is “metabolic diseases, nutritional deficiency states, and disorders due to toxins, drugs, and physical agents.” This topic encompasses many “must not miss” diagnoses when approaching a wide variety of chief complaints, and is worthy of individualized coverage in e-learning.

Our findings have several key implications. First, in any discipline, identified coverage gaps may help podcast developers identify topics for future episodes. Studies on medical learners’ listening habits have found that a subset of listeners seeks specific content corresponding with their areas of weakness [20]. This phenomenon is in alignment with the emerging concept of “microlearning”—with increasingly easy access to technology, learners may choose to acquire information and skills in small units to address gaps in their knowledge base [32]. As such, a library of available education content for any specialty would help learners access specific information in answer to clinical questions.

Second, this type of characterization may help educators identify podcasts or specific episodes to recommend to learners, or to integrate them into existing curricula. The past decade has seen the emergence of a new role for medical educators as “curators” [33], expertly selecting high-quality e-learning resources most fitting for their learners rather than creating content themselves. This important role is often underrecognized; there is growing evidence that some educators have little understanding of how their learners engage with e-learning [21]. Educators may also engage in helping learners appraise e-learning tools [34]. In general, educators have an important role to play in curating e-learning resources with the best fit for their learners.

Finally, our findings underscore the missed opportunity for featuring non-clinical topics in education podcasts. Experts have opined that such content, for example, on the hidden curriculum or professionalism, may be particularly well-suited for the audio format [35]. If podcast developers have little personal expertise on these topics, it may be important for them to feature known experts in their fields.

In terms of limitations, we selected our search terms to identify neurology education podcasts targeting a general neurology education audience, which are by definition imperfect given the multitude of ways in which trainees may search for this content. The titles of episodes may not reflect the entirety of their content, e.g., an episode titled “Clinical Manifestations of Parkinson’s Disease” may prominently feature content on associated neurobehavioral symptoms. Although we used a classification system for podcast topics developed a priori, there is also inherent subjectivity in classifications. As well, a large proportion of individual episodes were considered unclassifiable into the ACGME and/or ABPN categories, which affects the denominator used to calculate the proportion of content dedicated to different topics. Finally, the purpose of our study was to characterize availability of content and not on their use; in any case, statistics on number of downloads of individual podcasts are not generally publicly available.

In terms of future work, our study focused on characterizing content without examining the specifics of how learners consume it. One area of necessary study is learning analytics, defined as the examination and interpretation of meaningful patterns in data derived from learners in an education context [36]. This may include examining number of downloads (and trends in downloads over time), ratings on podcast platforms, and number of subscribers to associated social media platforms such as podcast-associated Twitter accounts (and trends over time) [37]. Learning analytics have been used to understand the optimal length of new media as defined by learner attention [38] and may be helpful for understanding the impact of the coronavirus 2019 pandemic on podcast listening habits [39]. Other future work may focus on evaluating podcast quality, which is inherently challenging given the subjective nature of existing rating systems. One study found that the gestalt rating of podcast quality requires approximately 20 health professionals to ensure interrater reliability [40]. Given the unregulated nature of podcast production and dissemination, academic societies and governing education bodies may invest in developing and validating peer review and objective scoring systems to evaluate podcasts [41].

Conclusions

Our study characterized the landscape of available education content in one medical specialty, which may help podcast developers fill gaps in coverage and help educators understand the content with which their learners may be engaged. The recent rapid expansion of available podcasts has outpaced our understanding of the format. As such, efforts to characterize content availability for individual disciplines are both warranted and timely.

Supplementary Information

Below is the link to the electronic supplementary material.

Data Availability

The datasets generated and analysed during the current study are available from the corresponding author on reasonable request.

Declarations

Conflict of Interest

The authors declare no competing interests.