The development of precision medicine has led to increased coverage in the US news media (Marcon et al., 2018; Ratcliff, 2021). News media are a powerful tool for reaching a wide range of audiences with information about new technologies (Caulfield and Condit, 2012). Also, news media connect the public with the scientific community and cultivate trust by providing readily accessible definitions and interpretations of scientific findings (Caulfield, 2018). As such, news media can become an important channel for the lay public to learn PM, exerting a strong influence on their awareness, attitudes, and engagement in related research and services (Woodbury et al., 2020; Ma and Kannampallil, 2021).

However, in explaining scientific breakthroughs to a broader audience, news media often create a gap in the information accessible to the public. The public often finds media coverage of scientific research too complex to comprehend (Morosoli et al., 2024). Also, news media do not always provide factual and balanced information related to new scientific technologies. According to the framing theory, journalists routinely frame a topic by reporting selected aspects of a topic and promoting “a particular definition, causal interpretation, moral evaluation, and/or treatment recommendation” (Entman, 1993, P. 52). Frames shape public understanding by organizing different issues and ideas of a complex issue and selectively assign prominence and significance to certain aspects (Dobmeier et al., 2023). In addition, media coverage prioritizes scientific events with more sensational values and controversies, leading to inflated expectations of preliminary findings and neglecting crucial uncertainties (Ratcliff, 2021).

Consequently, a lack of accessible, accurate, and balanced information erodes public understanding and trust in scientific advancements. One-sided representations of PM in the news can generate unrealistic expectations among an uninformed public, leading to misconceptions of PM research and misinformed decisions (Benjaminy et al., 2015; Hicks-Courant et al., 2021). Under-reporting of key information may lower the significance of PM to diverse populations (Fisher et al., 2019). As a result, PM could not effectively engage different population groups in the research and implementation processes to achieve health equity (Adams and Petersen, 2016; Fisher et al., 2019). Taken together, investigating media portrayal of PM through the perspective of framing can shed light on the way news media change how the public perceives PM with findings that inform education and outreach efforts (Brossard, 2013; Hicks-Courant et al., 2021).

This study aims to examine how news articles in US print media portrayed PM between 2015 and 2021. The patterns, details, and context of news articles about PM revealed in our study will pave the way for future investigations into how news media shapes public perception of and engagement with PM. While past research directly examining media representation of PM is limited, studies of similar scientific topics, including genetic testing (Chavez-Yenter et al., 2023; Dobmeier et al., 2023), genomics research (Hendy, 2024; Morosoli et al., 2024), and personalized medicine for cancer (Hicks-Courant et al., 2021) reveal recurring frames and characteristics in media coverage. First, when reporting PM and related genomic science, news media often highlight the potential to improve medical care but may not fully convey drawbacks (Marcon et al., 2018; Ratcliff, 2021). News media tend to emphasize the promises of new findings, highlight successful cases, and expedite the delivery timeline but ignore scientific uncertainties (Garrison et al., 2019; Ratcliff, 2021). Another prominent problem is the lack of clear communication on key topics, issues, and stakeholders that are relevant to diverse participants. News media do not fully disclose critical issues involving genetic and genomic research, such as safety, privacy, and costs (Basch et al., 2023; Chavez-Yenter et al., 2023; Morosoli et al., 2024). Only a small percentage of news articles covering genetic services used sources from healthcare professionals (Basch et al., 2023; Chavez-Yenter et al., 2023).

Further research is necessary due to the scarcity of research specifically assessing the representation of PM in US media. There is one previous study by Marcon et al. (2018) that assessed “personalized medicine” covered by North American print and online news stories. The present research is different from Marcon et al. (2018) in three ways.

First, Marcon et al. (2018) sampled and combined news articles covering “personalized medicine” and “precision medicine” whereas the present study specifically focuses on precision medicine in a US context. Despite the similarity, precision medicine and personalized medicine should be distinguished by different origins and emphases (Juengst et al., 2016). Precision medicine is a more specific research approach focusing on disease taxonomy based on synthesizing the variability of individual patients (Pokorska-Bocci et al., 2014; Kosorok and Laber, 2019). In contrast, personalized medicine is a more ambiguous term that encompasses both biological research and holistic improvement in the health care system and sociopolitical contexts for individualized health care (Pokorska-Bocci et al., 2014; Juengst et al., 2016). Thus, a specific technique of personalized medicine, such as tailored health services (Ko et al., 2016), might not be considered precision medicine by providers and patients in US health care. Thus, the media coverage of precision medicine should be examined separately from that of personalized medicine (Juengst et al., 2016).

Second, the samples of the present study and Marcon et al. (2018) were from different timeframes. Marcon et al. (2018) included news articles published before March 2016. However, the present study will sample most of the news articles published in and after 2016, which is a critical period to capture the news coverage of precision medicine as an emergent approach to health care in the US. After the launch of PMI in 2015, US scientists and policymakers formally endorsed the research approach of precision medicine, rather than personalized medicine (National Research Council, 2011; Jameson and Longo, 2015; Erikainen and Chan, 2019). Since then, precision medicine research has generated notable findings and attracted public attention (Woodbury et al., 2020; Ma and Kannampallil, 2021). As evidenced by Google Trends, “precision medicine” became the dominant terminology in the US after 2015 whereas searches for “personalized medicine” have declined (Google Trends, 2004–2024).

Third, the present study will examine more characteristics of precision medicine in news articles than Marcon et al. (2018). Certain scientific details, such as the causes of a disease, and conventional health care, were not assessed by the previous study. We will also evaluate non-scientific framing details that suggest nuances in themes, stakeholders, and interpretations (Peters and Dunwoody, 2016). The findings will help understand how news media influence public understanding and opinion of PM.

The present study will sample news articles related to PM from US regional and national print media from 2015 when PMI was launched until 2021. Through a content analysis, we will describe key characteristics and contextual issues of PM emphasized in the news, and also the theme, valence, and source used to present the information. The coded variables include three domains of factors used to report PM, including (1) target diseases and their related characteristic (e.g., affected populations, existing treatment), (2) characteristics of PM, and (3) non-scientific news frames (e.g., themes and spokespersons). Descriptive statistics and chi-square analyses will be used to describe the frequency and relationships among three domains of factors. The study will answer the following research questions.

The present study conducted a quantitative content analysis of news articles in US print media. Content analysis is a systematic research method that synthesizes text data of recorded communication into quantitative forms for assessment, replication, and comparison (Lombard et al., 2002; Krippendorff, 2012). Each news article was a unit of analysis.

News articles were sampled from ProQuest Global Newsstream, one of the largest databases of news sources (ProQuest, 2022). We provide two reasons that support the sampling of news articles from print media. First, news on print media provides more factual and contextual information to examine complex issues, like PM, through a thematic scope than television news (Eveland et al., 2002). In contrast, news from television media often focuses on episodic content of single events and pursues sensational values (Iyengar, 1994). Second, print media can provide content with more depth and breadth than fully online sites. By offering both offline and online access, print media expand its reach to a large news audience across population groups (Maier, 2010; Fletcher and Nielsen, 2017). News articles published by both regional and national print media were included because multiple modes of media access blur the boundary of media markets (Chyi et al., 2010). Moreover, we sampled articles from various sections in English-language, general-interest media, aiming to capture how PM is portrayed for the general public in the US. For this reason, non-English outlets or specialized interest media (e.g., magazine) that appeal to a focused audience were not optimal sources for this study.

The sampling time frame started from January 1, 2015, until June 1, 2021. The starting date was determined based on the announcement of the national PMI in the US by the Obama Administration (Collins and Varmus, 2015). The sampling keyword “precision medicine,” was used. Similar terms, such as “personalized medicine,” were not used for two reasons. First, the US formally defined precision medicine as an approach to health care that should be distinguished from personalized medicine (National Research Council, 2011; Collins and Varmus, 2015). Thus, using “personalized medicine” or other similar terms could include news articles about medical approaches not considered precision medicine in the US (Juengst et al., 2016; Ko et al., 2016). Second, Google Trends shows that compared with “precision medicine,” the search for “personalized medicine” has dwindled after 2015 (Google Trends, 2004–2024).

The database search initially returned 710 results containing the keyword “precision medicine.” The screening process removed duplicates and included news articles that were published in US print news media, and provided at least one sentence about PM and related characteristics. The final sample included 198 articles (N = 198) published by 31 US news media. The names of the news media are listed in the Supplementary Material.

The sampled articles were manually coded by three coauthors as the coders. We randomly selected 20% of the full sample (n = 41) to train coders. The coder training reviewed and clarified the conceptual and operational definitions of the coded variables. Three coders coded a small batch of the training articles independently and convened again to discuss any disagreements. The coding scheme was adjusted to address coder disagreement. This training process was repeated until all of the training articles were coded and the intercoder reliability reached a satisfactory level (Lombard et al., 2002). Given substantial imbalances of frequencies in different categories, we used Gwet’s AC₁ to assess intercoder reliability. Gwet’s AC₁ offers an optimal statistical solution to overcome the paradox of high degrees of agreement and low intercoder reliability scores that plague other indices (Gwet, 2014). Gwet’s AC₁ intercoder reliability coefficients were computed and qualified by irrCAC in R (Gwet, 2022). The AC₁ coefficients of all variables had significant probabilities (>95%) that could reach the “substantial” or “good” level of agreement on the benchmark scales by Wongpakaran et al. (2013). All coefficients are included in the Supplementary Material. After achieving intercoder reliability, the training coding results of one coder were randomly selected and incorporated into the final results. Cumulatively, the training time was approximately 40 h.

After the training, three coders equally split the remaining sample and each manually coded one-third of the articles. After half of these articles were coded, 10% of the articles in each coder’s part were randomly selected and then cross-coded by two other coders to assess reliability. The coders convened again to solve any questions and disagreements. The coders established satisfactory intercoder reliability in the first half of the sample before starting to code the second half of the sample.

The initial coding scheme was developed by reviewing medical and public health literature on PM research (see references in Supplementary Material), and further adjusted during coder training through inductive, open coding of a small sample. The coding scheme included three domains of variables related to the media representation of PM. The first group, characteristics of PM, includes benefits and concerns of PM, the timeframe of achieving benefits, and the present status of PM. The second group of variables was target diseases and the related factors, which included the type and name of diseases, the primary cause of the disease, the affected population, and coverage of conventional health care. The last group, non-scientific news frames, includes meta-data of a news article, main theme, framing valence, and spokespersons. The Supplementary Material provides the operational definitions and coding instructions.

The study reported descriptive statistics to answer RQ1 and RQ2. Chi-square (χ²) statistics were used to examine the relationships between categorical variables related to the characteristics of PM, target diseases, and non-scientific news frames.

About 58.59% of the sampled articles were published in 2015 and 2016. The main theme of 151 news articles was related to PM’s function and benefits in health care, which includes treatment, prevention, and diagnosis (see Table 1). Fewer articles had a theme related to the risks and concerns of PM. Other themes were not frequently identified in the sampled articles.

RQ1 was to describe the characteristics of PM portrayed in news articles. The analyses revealed that 92.42% of the news articles reported at least one benefit of PM. The most common benefit was improved treatment efficacy, which was followed by disease prevention (Table 2). Relatively fewer cases reported reduced adverse side effects, lower cost, and advancing science as the benefits.

Table 3 summarizes the concerns of PM described by news articles. Only 61 news articles (30.81%) discussed the concerns or risks of PM. Insurance and policy restrictions and efficacy of PM were the two most frequently discussed concerns.

Most of the articles did not provide a specific time frame for which PM can deliver its benefits. About 34.85% indicated the benefits of PM were expected to be realized in the future and another 14.65% suggested the benefits to be achieved soon. In addition, 62.63% of the news articles described PM as ongoing research whereas 19.19% portrayed PM-related products as being manufactured or readily available. Fewer articles (6.06%) described PM as a new concept.

RQ2 was to identify diseases and their related factors (i.e., disease causes, non-PM conventional health care, and affected populations) in news articles reporting PM. The analysis found that 61.62% of the articles (n = 122) mentioned cancer, making it the most common disease covered in association with PM. About 10.61% of the articles (n = 21) related to other non-cancer chronic diseases. Cystic fibrosis, heart disease, and diabetes were the three most common non-cancer chronic diseases. Rare genetic disorders and mental illness were found only in 6 (3.03%) and 7 (3.54%) of the articles. All three articles (1.52%) related to acute or infectious diseases were about the relationship of COVID-19 with PM. Another 15.66% did not focus on any specific disease or health condition.

In addition, about 20.71% of the news articles mentioned non-PM conventional health care. Genetic or hereditary factors were portrayed as the causes of diseases in 35.86% of the news articles (n = 71) but the most of the news articles (62.63%) did not provide a cause.

The majority of the news articles (n = 166, 83.84%) did not specify the populations affected by the disease who could potentially benefit from PM. There were 12 cases (6.06%) discussing one particular gender (women or men), and 7 (3.54%) focusing on individuals of certain genetic or family history factors.

RQ3 sought to understand how news articles portrayed diseases and related factors when reporting the characteristics of PM. Multiple chi-square analyses were conducted to answer RQ3. The first set of analyses was to investigate the diseases linked to specific benefits or concerns of PM in news articles. The results showed that cancer was discussed in 64.50% of the news articles that indicated the improved efficacy as the benefit of PM. By contrast, 8.88, 3.55, and 3.55% of the articles featuring this benefit were, respectively, related to non-cancer chronic diseases, genetic conditions, and mental illness, χ² (6, N = 198) = 12.86, p = 0.045, V = 0.26. Most cases (62.50%) pointing to economic development as the benefit of PM did not specify any health conditions, χ² (6, N = 198) = 14.37, p = 0.03, V = 0.27. On the other hand, the coverage of the concerns related to PM did not specifically focus on one type of disease (ps > 0.05).

Second, we investigated what benefits and concerns of PM were mentioned alongside discussions of non-PM conventional health care in news articles. The chi-square analyses showed that non-PM health care was mentioned in 83.33% of the news articles framing fewer adverse side effects as a benefit of PM [χ² (1, N = 198) = 14.78, p < 0.001, V = 0.27]. Discussions of non-PM health care also appeared in about 66.67% of the articles that framed increased patient autonomy and life quality as the benefit from PM [χ² (1, N = 198) = 7.96, p = 0.005, V = 0.20]. Most of the news articles (57.89%) discussing PM’s potential for reducing health care costs failed to discuss existing treatments, χ² (1, N = 198) = 5.86, p = 0.02, V = 0.17.

Third, benefits and concerns were reported using different time frames. About 36.69% of news articles portrayed the improved treatment efficacy of PM as something to be achieved in the future χ² (2, N = 198) = 10.42, p = 0.01, V = 0.23. A smaller percentage of such articles (17.16%) highlighted immediate availability of improved efficacy from PM. By contrast, most of the concerns regarding PM were not tied to a specific time frame in news articles (ps > 0.05), acknowledging the uncertainty of addressing these challenging issues.

RQ4 was proposed to identify what non-scientific news frames, including main theme, framing valence, and spokespersons, were employed in news articles reporting the characteristics of PM. To answer RQ4, one of the analyses was to examine different types of spokespersons in the news articles about PM. The descriptive analyses show that research scientists were the most cited spokesperson, followed by biotechnology companies, patients or patient groups, and politicians or government officials (Table 4). Other sources were less frequently cited as spokespersons. The chi-square analysis show that scientists were mostly cited in the news articles with a theme about PM’s function and benefits in health care (78.62%) as well as those about risk or safety concerns (11.03%), χ² (6, N = 198) = 16.70, p < 0.05, V = 0.29.

To understand the relationship between spokesperson and the present status of PM research, the chi-square analysis found that patient or patient groups were cited in 39.47% of the news articles that framed PM to be a product or service being manufactured or already available, χ² (3, N = 196) = 11.05, p = 0.01, V = 0.24. Only 2.63% of the articles describing PM as an available service cited politicians or government officials, χ² (3, N = 196) = 11.19, p = 0.01, V = 0.24. By contrast, when portraying PM as a newly conceptualized idea, only 8.33% of the news articles cited patients but 41.67% cited politicians or government officials. When framing PM as an ongoing research project, the proportion of the articles citing politicians or government officials (20.97%) was slightly higher than that of the articles citing patients (16.13%). Scientists and other spokespersons were not cited differently in the articles discussing the status of PM research (p > 0.05).

Lastly, we examined the valence of news articles when reporting the benefits and concerns regarding PM. The chi-square analyses indicated that the vast majority (81.07%) of the news articles addressing treatment benefits of PM expressed a positive tone, χ² (2, N = 198) = 29.33, p < 0.001, V = 0.39. In contrast, about 40.9 and 50% of the news article expressed a negative tone when discussing the concern of PM related to limited efficacy [χ² (2, N = 198) = 47.50, p < 0.001, V = 0.49] or difficulty of conducting research [χ² (2, N = 198) = 27.16, p < 0.001, V = 0.37], respectively. The proportions of the articles showing a positive tone when reporting these two concerns were lower (22.73 and 20%, respectively). The valence was more mixed in the discussion of legal, policy, and insurance-related concerns. Each framing valence was found in respective one-third of the news articles focusing on these concerns, χ² (2, N = 198) = 40.68, p < 0.001, V = 0.45.