Reading, writing and research impact

Yesterday two things happened that crystallised for me what we are doing here in publishing the Journal of Primary Health Care. The first thing was my lecture to third year undergraduate medical students about research. They are so keen to ‘do research’, without knowing the first thing about how or why it is done. They so desperately want to change the world right now, as soon as possible, through their amazing research (topic unknown). My view is that their main job as undergraduate medical students is to learn about clinical medicine so that they can become good doctors, and that reading the scientific literature is a critical part of their job, both as students and later as they practise medicine. No one had told them much about reading anything other than lecture notes or medical textbooks so it was a new idea to read beyond this limited literature, and to read as a professional responsibility rather than to pass exams.

The second thing happened while I was helping a friend and colleague to update his curriculum vitae. He is now a full-time medical academic but for decades he was a rural general practitioner, burning candles at both ends while fitting teaching and research around his patients’ needs. We came to a paper he had co-authored and he mentioned that according to ResearchGate it had 2500 ‘reads’ in the 18 months since publication.¹ My eyebrows shot up in amazement as (my bad!) I do not typically think of him as a high impact researcher and that statistic shows that this is a really well read paper.

‘Impact’ is a concern for scientific journal editors, closely following our primary concern, which is to publish articles that our audience likes to read. The big problem with these performance tasks is that there is no very robust measure of them. This is especially the case for editors of journals like the Journal of Primary Health Care, that are specifically designed to serve clinical audiences. Laboratory-based scientists use citations as their measure of impact and the notion that citations should be the main concern of scientific researchers has spread to the applied medical disciplines. Effectively, counting citations means that ‘impact’ is measured by counting the extent to which one person’s research is built on in another person’s research. Citations are not a good measure of impact in the clinical sciences because our primary concern is to not to influence other scientists’ research but to publish research that might influence the way clinicians care for patients or how policy-makers construct health systems. For us, reading is the ultimate impact measure because it is only after a paper is read that it has a chance of influencing patient care.

The Journal of Primary Health Care has recently attached an ‘Altmetrics’ impact measure to each paper.² As the name suggests, Altmetrics is an alternative measure of research impact that uses social media (such as Facebook, Twitter, and news outlets) to measure the amount of attention a paper attracts. We have published several papers that have very high Altmetric scores and that makes us happy (eg^3–6). However, Altmetrics does not tell us about the ultimate measure (reading) and there is not much association between an Altmetrics score, or the number of times a paper is cited, or the number of times a paper is read. For example, even though my friend’s paper had 2500 reads, its Altmetric score is 6 and it has 4 recorded citations. Reading is very hard to measure and, like every other measure of science impact, should be taken with a grain of salt. There is no direct measure yet of the impact of scientific publications on patient care.

Even so, this issue of the Journal is full of research for readers to think about. Our lead article investigates the receipt of nutritional advice by 16 people with various long-term conditions, revealing (again) that patients find it hard to maintain healthy diets.⁷ However, this article also raises the suggestion that maybe healthcare providers need to think a little differently about how to advise their patients. It suggests that considering the Whare Tapa Wha model of health⁸ when delivering dietary advice might positively influence its uptake. Along similar lines, Rolston’s team report research showing gaps in the knowledge about gout of both Pakeha and Maori patients that, if filled, would probably improve their health.⁹ We learn that sometimes it is better to talk to people than hand out pamphlets to assist patients with gout to understand their illness, but sometimes the opposite applies. We have three more research articles about clinical topics. Murray and Roke use data from family planning clinics to define ethnic differences in the use of different contraceptives that seem to rest on financial grounds,¹⁰ and we have two papers about melanoma. New Zealand has a higher prevalence of skin cancer than any other country in the world and Townsend shows what this means for the organisation of care in his rural general practice.^11,12 Brian and Jameson include the hospital side of the melanoma story, concluding that New Zealand guidelines for melanoma excision biopsy need revision so that they provide more clinically meaningful measures that clinicians can use to audit their care.¹³ Further relating to the practical use of guidelines, Reyneke, Jaye and Stokes examine the process of accepting and using care pathways in the Western Bay of Plenty.¹⁴

We also have an array of papers about health care services. Nixon shows how point of care ultrasounds in rural hospitals influence generalist physicians’ decision-making.¹⁵ Soh and Low reviewed the literature on house calls by general practitioners and in doing so they have inadvertently set an agenda for research in this issue of increasing importance as populations age.¹⁶ Why do we not know more about the epidemiology of house calls – who is visited, why and where? Is the call for house calls going to become more pressing? If so, when? How much do house calls cost and who pays?

Szafran’s team from Canada show how patients value the team-based care that has been implemented in Canadian primary care clinics for more than a decade¹⁷ and McKinlay’s team extend our knowledge about the importance of social networks by testing the knowledge in general practice about patients’ social networks by the statements of patients themselves of their own networks.¹⁸ Strobel et al. use routinely collected data to examine process of care indicators for indigenous children in Australia.¹⁹

Finally, check out our columnists discussions of saline eyewashes and echinacea might give you some useful practical tips for accessible treatments for common conditions.^20,21 Enjoy your reading!