No such thing as bad publicity? A quantitative content analysis of print media representations of primary care out-of-hours services

OBJECTIVE: To explore how out-of-hours primary healthcare services (OOHS) are represented in UK national newspapers, focusing on content and tone of reporting and the use of personal narratives to frame stories. DESIGN: A retrospective cross-sectional quantitative content analysis of articles published in 2005, 2010 and 2015. DATA SOURCES: Nexis database used to search 10 UK national newspapers covering quality, middle-market and tabloid publications. INCLUSION/EXCLUSION CRITERIA: All articles containing the terms ‘out-of-hours’ (≥3 mentions per article) or (‘NHS 24’ OR ‘NHS 111’ OR ‘NHS Direct’) AND ‘out-of-hours’ (≥1 mention per article) were included. Letters, duplicate news items, opinion pieces and articles without a substantial portion of the story (>50% of an article’s word count, as judged by researchers) concerning OOHS were excluded. RESULTS: 332 newspaper articles were identified: 113 in 2005 (34.1%), 140 in 2010 (42.2%) and 79 in 2015 (23.8%). Of these, 195 (58.7%) were in quality newspapers, 99 (29.8%) in middle-market and 38 (11.3%) in tabloids. The most commonly reported themes were OOHS organisation, personal narratives and telephone triage. Stories about service-level crises and personal tragedy, including unsafe doctors and missed or delayed identification of rare conditions, predominated. The majority of articles (252, 75.9%) were negative in tone. This was observed for all included newspapers and by publication genre; middle-market newspapers had the highest percentage of negative articles (Pearson χ(2)=35.72, p<0.001). Articles presented little supporting contextual information, such as call rates per annum, or advice on how to access OOHS. CONCLUSION: In this first reported analysis of UK national newspaper coverage of OOHS, media representation is generally negative in tone, with frequent reports of ‘negative exemplars’ of OOHS crises and fatal individual patient cases with little or no contextualisation. We present recommendations for the future reporting of OOHS, which could apply to the reporting of healthcare services more generally.