The definition of ‘health-literacy’ that was developed for the National Library of Medicine and used by Healthy People 2010 [17] is as follows:
“The degree to which individuals have the capacity to obtain, process, and understand basic health information and services needed to make appropriate health decisions”[18]
It has been suggested that inadequate levels of health literacy have a negative effect on patient health outcomes, and can be a stronger predictor of patient health status than age, income, employment status, education level, and racial or ethnic group [19]. Online health literacy can be improved in 2 ways: (i) improving the health literacy of the population, or (ii) decreasing the health literacy levels needed to read and understand online health information. This study has shown that the reading difficulty of most online information far exceeds the levels to which the average patient may be able to comprehend. The mean FRE was 38.86 ± 13.27 which is at a college level reading standard, while the mean GF was 12.11 ± 2.88 and FKG was 10.59 ± 2.15 which are at a 12th and 11th grade level respectively. This is far from the 6th grade reading level recommended by the American medical association (AMA) [20] and agency for health research and quality (AHRQ) [21], in fact none of the URLs meet this criterion when analysing FRE scores and only 4 and 2 were under this level according to GF and FKG score respectively. This would make the information quite inaccessible for a lot of patients, with most URLs even failing to pitch this information below the US reading level average (8th grade), as 6 papers were under this with FRG scores, 4 with GF, and 3 with FRE.
The HONcode certification aims to provide the reader with trustworthy health related information, in this study less than 9% of the URLs analysis had HONcode certification. While the HONcode certification does not necessarily guarantee high quality information, a greater percentage of those with HONcode certification were found to uphold the principles set out by HON (89% vs. 0.6% compliance [16]). Such a low number of certified websites in this analysis means the majority of online information the patient is consuming can be assumed to be of poor quality and is likely untrustworthy. Online-health information has the potential to benefit patient consultations and lead to better outcomes, however the issue arises when the information the patient is consuming is false or misleading. This ultimately leads to confusion, distress, and a tendency for patients to harmful self-diagnosis and/or self-treatment [22]
The JAMA score is a means of analysing the quality of websites manually. This tool was used as URLs could be analysed for quality regardless of application for HONcode certification. Approximately one quarter of websites satisfied all 4 criteria, while approximately one fifth satisfied no criteria. This shows the information patients are reading may not be of good quality and may lead to breakdown in patient-physician communication [22]. JAMA benchmark criteria mainly assess the quality of online information by assessing factors associated with the formality of information [23]. A limitation of using JAMA criteria for analysis of quality is that it requires the consumer to use several indicators to come to a judgment, this is not practical for a patient and requires prior analysis of the URL. Development of an automated JAMA scoring tool would be helpful for patients and physicians to assess URLs for quality.
The Rockwood classification system has helped to guide management of AC joint injuries however there is still debate regarding operative vs. non-operative management of grade III injuries. It has been found that there is poor correlation between Rockwood classification and pain (rs= 0.05; p = 0.752) and also poor correlation between Rockwood classification and functional deficit [24]. A meta-analysis on type III injuries found no difference between non-operative treatment and operative with regard to strength, pain, throwing ability and incidence of acromioclavicular joint osteoarthritis [25]. There was however found to be a better cosmetic appearance with operative management (p < 0.0001), while operative management had significantly greater duration of sick leave in comparison to the non-operative group (p < 0.001). This emphasizes the importance of individualizing patient management, especially when dealing with type III injuries. Upper limb surgeons across the UK were surveyed regarding treatment of type III injuries [26]. Some surgeons commented that they leave it to patients to decide. This shows the importance of informed patient decision making, thus avoid unwanted surgery and unwanted complications [27–28] with disruptions to the patients daily living [25]. The readability scores for the URLs reviewed are higher than recommended, this would make the information inaccessible to some patients and would therefore disrupt the shared decision making process. Health literacy levels must be accounted for when designing a patient decision aid (PtDA), with lower health literacy levels accounted for so that all patients can engage appropriately in the shared decision making process [29].
There are some limitations associated with this study. General literacy levels are not synonymous with health literacy levels, however measurements of general literacy levels have been used as a surrogate to estimate health literacy levels [19]. Health literature can be littered with common medical abbreviations but are unfamiliar to the lay person. This may have resulted in certain sites receiving artificially lower scores as the readability scores are calculated using words, syllables and sentences as parameters. The search was completed by the lead author in their home country on their own device laptop and thus search results could differ. Lastly, videos and visual aids were not part of the analysis as it would be difficult to assess for reading level and quality. However, visual aids and videos would greatly enhance a patients understanding and comprehension of a topic.
The strengths of this study are that it has provided a comprehensive review of currently available online information on acromioclavicular joint reconstruction. There is a vast amount of information available online regarding this topic, however, the majority of this information is not accessible to the general population as it is of poor quality and is difficult to read. The deficiencies in the current information can be used to help address these issues in future online articles.