Bibliometric and visualized analysis are regarded as appropriate tools for describing the present status and predicting the future trends concerning the research of interest. In this study, the current status and global trends of hip fracture research were delineated. The quantity of publications has gradually increased yearly. Leading researchers, contributing institutions, countries and their cooperation relationship have been identified, and important publications with high-citations highlighted.
Utilizing co-occurrence cluster analysis, we presented a network map of co-occurrence relationship by analyzing the keywords of all the included studies. Totally, five potential research orientations were identified and keywords as mortality, osteoporosis, risk, elderly-patients and etc. were highlighted with bigger icons. These results suggested that hip fractures, especially in older individuals, lead to conditions that extend far beyond the orthopaedic injury, with consequences in the aspects of epidemiology & preventive medicine, internal medicine & endocrinology, as well as critical care and gerontology. This is also confirmed in Fig. 2. Using the overlay visualization map of the co-occurrence analysis, different colors represented the relevant year of publication. Nodes of various colors (from blue to red) could all be found with substantial densities in the five clusters, which suggested a pattern of balanced development existed in these five investigation directions respectively during the past decade. Specifically, a pattern of balanced development may exist within the field of hip fracture, accompanied with inherent alterations of hotspots in each sub-orientation. Additionally, each direction itself was also experiencing the changes of research hotspot, suggesting a diversified developing trend. Also by using this network map together with other bibliometric information, we had chances to further shed light on the growing trends in the past, present and the future.
The number of citations amassed by a publication could be employed as a surrogate marker of the impact made within its field19. Interestingly, 90% of the leading 10 hip fracture studies with the largest overall citation count feature in the top 10 for annual citations, implying both historical significance as well as significance to current medical practice. Most of these highly-cited studies were published before 2010, as prior studies have longer time period of citation accumulation inherently. They mainly focused on osteoporosis management, fracture epidemiology and prevention, as well as perioperative management and patient safety.
I. Fracture prevention and medication therapy
The most-cited report was published by Dr. McClung in 2001 describing the protective effects of risedronate which substantially minimized hip fracture risk among aged females with established osteoporosis but not among those with risk factors apart from lower BMD, i.e. poor gait or fall propensity16. From this large trial (9331 female patients), the results demonstrated the value of BMD measurements in identifying women for whom medication therapy to prevent hip fracture is adequate. Similarly, a double-blinded randomized controlled trial (RCT) discovered that an annual infusion of zoledronic acid within 90 days following surgical fixation of a fragility hip fracture was associated with a reduction of a new clinical fracture rate together with improved survival17. While conversely, a nested case-control study carried out utilizing the General Practice Research Database (1987–2003) in UK discovered that long-term proton pump inhibitors therapy, especially at high doses, was associated with an elevated hip fracture risk20.
On fracture risk: Other research centered on hip fracture epidemiology. In the United States, hip fracture rates and consequent mortality among individuals 65 years and older were decreasing along with the usage of bisphosphonates and comorbidities among patients with hip fractures had increased21. Hip fracture risk could be predicted by BMD and clinical risk factors(CRFs). The prediction model, with the combined use of CRFs and BMD than BMD alone, could be improved with a greater gradient of risk (risk ratio/standard deviation change in risk score) from 3.7/SD to 4.2/SD22. An additional systematic review(SR) reported that age-standardised rates of hip fracture were accessible for 63 countries23. Additionally, there was a greater than 10-fold variation in hip fracture risks between nations. High risk regions for men were Taiwan, Austria, USA (Caucasian), Switzerland, Norway, Sweden and Denmark. Those at low risk included Tunisia, Oceania, the Latin American countries of Ecuador and Colombia and several European countries (Spain, Poland, Romania, France and Turkey). Other countries at low risk were China, Lebanon, Philippines and the US Black population. The basic pattern of fracture likelihood in women was comparable to that in men. And discordances within classification were fairly few. Five countries coded as low risk in men were at intermediate risk for women (Poland, New Zealand, Romania, France and Turkey). Seven countries coded as moderate risk in men were coded at high risk in women (Japan, Belgium, Singapore, Canada, Malta, UK and Slovakia)23.
On mortality: a meta-analysis found that older adults have a 5- to 8-fold higher risk for all-cause mortality throughout the first 3 months after hip fracture. Excessive annual mortality persisted for both women and men even after 10 years of follow up (FU), but at any given age, excessive annual mortality after hip fracture was higher in males than in females24.
III. Perioperative management and orthogeriatric collaborated care
A prospective observational cohort study reported that in elderly patients, the existence of three or more comorbidities would be the strongest preoperative risk factor for mortality within the 1st month after surgical procedures. Chest infection and heart failure were the most common early postoperative complications and resulted in increased mortality. These groups offer an apparent target for specialized medical evaluation25. Another RCT noted that proactive geriatrics consultation was effectively applied with good adherence after surgical procedures. It diminished delirium by over one-third, and reduced severe ones by over one-half. The trial provided strong preliminary evidence that proactive geriatrics consultation played a crucial role within the acute hospital management for hip fracture patients26.
8 of these 10 highly-cited articles were published from 2000–2009. While studies published from 2010 to 2014, continued to put attention on patient safety and mortality, but in an extended time point of view and FU. A SR reported that the overall inpatient or 1 month mortality was 13.3%, 3–6 months mortality was 15.8%, 1 year 24.5% and 2 years 34.5%27. There were strong evidence for 12 predictors, including advanced age, male gender, nursing home or facility residence, poor preoperative walking capacity, poor activities of daily living, greater American Society of Anaesthesiologists grading, multiple comorbidities, diabetes, cancer, cardiac diseases, poor psychological status, dementia and cognitive impairment27. A population-based study discovered that in 4 years of FU, the most common causes of death were circulatory diseases, followed by dementia and Alzheimer's disease28. Besides, in this 5 years (2010–2014), researcher started put focus on surgery timing as well as orthogeriatric care and their impact on patient outcomes. Earlier surgery (less than 4829 or 72 hours30) was associated with a reduced risk of death and lower rates of postoperative pneumonia and pressure sores among aged patients suggesting minimizing delay to improve outcomes30. A meta-analysis supported that ortho-geriatric cooperation to improve mortality after hip fracture fixation31. Immediate admission of patients aged 70 years or more to comprehensive geriatric care in a dedicated ward improved mobility at 4 months, compared with the usual orthopaedic care alone32.
IV. Patient rehabilitation, medical and economic burden
Since 2015, research began to concentrate on patient rehabilitation, economic burden of the injury as well as national audit and registry studies. The fracture has a considerable impact on older peoples' medium- to longer-term capabilities, physical functions, quality of life and accommodation. Only 40% to 60% of study participants recovered their pre-fracture level of mobility and ability, while 40%-70% regained their level of independence for fundamental activities of daily living. For individuals independent in self-care prior to fracture, 20%-60% needed support for various tasks at 1 and 2 years FU 32. These results suggested that great medical and potential economic burden existed for hip fracture survivors. Medical expenses following hip fracture were high and mostly occurred in the first year after the index fracture, while suffering from a second hip fracture accounted for a great deal of the expense increase. There is a solid economic incentive to prioritise research funds towards figuring out the best strategies to prevent both index and subsequent hip fractures33.
V. National audit and initiative
The great burden caused by hip fracture warranted actions or projects in a greater scale, i.e. in country level. The UK National Hip fracture Database was launched in 2007 being a national collaborative, clinician-led audit initiative to enhance the hip fracture care quality, which was associated with significant improvements in care and survival of aged individuals with hip fracture in England. From 2007 to 2011, early surgery rate increased from 54.5% to 71.3%, which had remained stable from 2003–2007. Thirty-day mortality fell from 10.9% to 8.5%, in comparison with a smaller reduction from 11.5% to 10.9% previously. The yearly relative decrease in adjusted 30-day mortality was 1.8% per year in the time period 2003–2007, compared with 7.6% per year over 2007-201134.
A clear developing trend in hip fracture research in the past 2 decades could be described, which initiated from fracture epidemiology and prevention, transitioned to perioperative management, orthogeriatric care and patient safety, and then to patient rehabilitation, disease burden as well as national audit and initiatives in recent time period.
As the number of publications continuously increases and significant burden of hip fracture prevails, an underlying vigorous development of hip fracture study could be expected with a growing quantity of papers published in the next couple of years. In addition, the co-occurrence analysis could depict the developing trends and research hotspots, which might further provide the investigators with inspirations of topic selection and assist the funding agencies make profit investment plans. Through the results of the overlay visualization map in co-occurrences analysis, "timing of surgery, registry and patient mobility” accounted for larger ratios for color yellow, indicating more studies dedicated to patient safety, functional recovery and big data research after 2010. Additionally we used the second data set from 2018 to 2020 in co-occurrence analysis and identified several key words, i.e. “surgery”, “outcomes”, “elderly-patients”, “management”, “complications” with increasing occurrences in recent years shown in Table 5. Given the above two similar and interrelated results, we anticipated several hotspots within hip fracture research. 1. Optimization of peri-operative management and complications prevention; 2. Post-injury rehabilitation and care; 3. Meta-analysis, registry and big data research.
This study inevitably has some limitations. Firstly, there are intrinsic differences between the results of bibliometric analysis and real-world study. For instance, some comparatively new publications with high quality may not attach sufficient attention due to lower citation frequency, while older articles have a tendency to accumulate more citations. A second limitation is the 'obliteration by incorporation' effect describing the bias created with citation analysis which occurs when particular ideas become so accepted that the most original work is no longer cited35. Also self-citing (or neglecting to cite a rival 's work) might bring in the inherent bias of' incomplete citing' and' omission bias.' Third, bias may still exist when considering the same short name or various expressions of certain authors and keywords. However in this study, with the usage of bibliometric and visualized analysis, hotspots and collaborative relationship among countries, authors, and institutions were identified, meanwhile scientific masterpieces highlighted and reviewed. These information could provide investigators a vivid general view within the academic field of hip fracture research. A time trend was depicted from its epidemiology, osteoporosis management and fracture prevention in the 1st decade of 21th century, to patient mortality and surgery timing in the later time period, to rehabilitation as well as national registry and audit research in the last period. These information could also guide stakeholders in priortising funding and optimizing care of hip fracture.