Abstract
Objective
The internet is increasingly being used as a resource for health-related information by the general public. We sought to establish the authorship, content, and accuracy of the information available online regarding platelet-rich plasma (PRP) therapy for knee osteoarthritis.
Design
Top 200 search results from each of the 3 leading search engines available online (Google, Yahoo!, Bing) were screened, and 181 websites were finally reviewed for content with emphasis on specific claims, comparing between websites authored by private physicians/groups and other authorship types.
Results
Nearly 80% of the websites claimed that PRP injections for osteoarthritis of the knee improve patients’ pain. A total of 42.8% of the private websites and 27.6% of nonprivate websites have stated that the procedure can delay or eliminate the need for future surgery. Costs were only mentioned by few (11.6%), and mainly by the nonprivate websites. Both website groups were unlikely to mention that PRP therapy is not the treatment of choice for end-stage knee osteoarthritis (7.9% of private and 17.2% of the nonprivate sites), or to state that patients with less advanced disease may benefit more from the treatment (11.8% and 20.6%, respectively). Private websites were less likely to refer to peer-reviewed literature (18.4% vs. 41.4%) and were more than 3 times less likely to mention lack of adequate evidence (13.2% vs. 48.2%).
Conclusions
Patients seeking online information regarding PRP therapy are vulnerable to websites presenting a narrow viewpoint of this treatment modality, putting emphasis on unsubstantiated benefits while disregarding potential drawbacks and concerns.
Introduction
The popularity of platelet-rich plasma (PRP) therapies has increased significantly over the past few years with an emphasis on orthopedic conditions.1,2 This rise in popularity could be attributed to the possibility of enhancing reparative processes in tissues with low healing potential. PRP is created through a centrifugation process called plasmapheresis, in which liquid and solid components of anticoagulated blood are separated. 3
Laboratory studies have shown that PRP treatment of osteoarthritis joint tissues result in decreased catabolism, suppression of inflammatory mediators, 4 and an increase in cartilage synthetic activity. 5 Various orthopedic applications have been trialed, including bone healing, muscle injuries and regeneration, tendinopathies, and osteoarthritis, aiming to manage pain by tissue regeneration in both acute and chronic clinical settings. 3 There is a growing amount of data on the benefits of PRP and knee osteoarthritis, in terms of pain relief and self-reported function improvement.6,7. However, other studies suggest that PRP does not yield significantly better outcomes in the long term 8 or when compared to hyaluronic acid (HA).9,10
The published literature surrounding the use of PRP varies greatly. Current guidelines from the American Academy of Orthopaedic Surgeons (AAOS) indicate an inability to “recommend for or against” the use of PRP for the treatment of knee osteoarthritis while citing that the literature had moderate applicability and strength of evidence.11,12 Likewise, the American Association of Hip and Knee Surgeons, the Hip Society, and the Knee Society have just recently published a position statement regarding the use of biologics for advanced hip and knee arthritis, stating that biologics, including PRP, cannot be currently recommended since the efficacy of these treatments is yet to be established while the potential for harm is likely underappreciated. 13
Several studies have shown that patients at different stages of osteoarthritis do not have the same response to the treatment. Filardo et al. 14 concluded that PRP was more effective in young patients with early or moderate stages of arthrosis, but it had a limited effect in cases of advanced osteoarthritis. Chang et al. 15 evaluated the effects of intra-articular injection of PRP or HA in a systematic review and reported better results among those patients with milder forms of osteoarthritis than advanced forms. Sanchez et al. 16 showed that PRP intra-articular injections in combination with intra-osseous infiltration in severe knee osteoarthritis were not effective and did not provide any benefit. 16
The internet is becoming a major source of health care information, with content that can range from being educational and professionally reviewed to personal blogs, opinions, or anecdotes of other patients. 17 According to a recent report by the Pew Internet Project, 80% of U.S. adult web users have searched for health-related information online. 18 A recent study estimated that approximately 40% of patients will turn to the internet before undergoing surgery. 19 As online health information content, major concerns arise about the accuracy and quality of the available information,20,21 as many websites operate without filtering by professional editorial review. 22 Previous studies suggested that much of the information online is imprecise, incomplete, or out of date, and although a large portion of the information is commercial, it cannot easily be identified as such by readers. 20 Without proper guidance, inaccurate health information on the internet can be harmful and misleading to patients 23 and potentially costly to health care systems and society as a whole. More specifically, orthopedic-related information available online has been shown to vary in accuracy, quality, and readability.24-27
The purpose of this study is to evaluate the content, and readability of online information regarding PRP for knee osteoarthritis, which is immediately available to internet users through the major popular search engines, within their first 200 results. We hypothesized that the information available online will include specific unsubstantiated claims in favor of PRP treatment, particularly when authored by private physicians or groups.
Methodology
Using a previously described methodology19,24,28 we used the 3 most popular search engines (Google, Yahoo!, Bing), to conduct a search for websites or webpages containing information regarding PRP. The exact search string was: (“PLATELET RICH PLASMA” OR “PRP”) AND” KNEE” AND (“ARTHRITIS” OR “OSTEOARTHRITIS”). The search was carried out on March 7, 2018.
The first 200 results from each search engine, Google, Yahoo! and Bing, were screened for eligibility. We excluded promoted (paid content) websites, websites presenting a wide overview on biologics and not solely on PRP, websites relating to PRP treatment for indications other than knee OA, results directing to videos not supplemented by text, and results for which the authorship was not stated. The initial search results were combined together, and duplicate results were filtered. A combined list of 244 websites was further narrowed down: 31 websites were excluded for providing information on combined treatment of PRP and other biologics and 9 results directing to videos not supplemented by text were also excluded. Twenty-two results were dismissed due to unsuitable etiology for PRP treatment (other than osteoarthritis) and an additional site was removed as the authorship was not stated ( Fig. 1 ).
Flowchart depicting the websites selection process.
The final data was composed of 181 websites that provided information specifically on PRP treatment. Of the final list of websites, 112 websites were found in the initial Google search (before duplicate filtering and screening), 107 in the Yahoo! search, and 95 in the Bing search. Links to other webpages and PDF (portable document format) files leading to information within the same website were followed and included in the analyses, as opposed to links leading to external websites which were omitted.
Eligible websites were then evaluated for authorship and content.
Authorship
Authorship was categorized for each eligible website. The categories included (1) hospital or university, which identified a group of physicians to an academic institution or hospital system; (2) private physician/clinic with no such affiliation; (3) orthopedic manufacturer/industry in the field of PRP; and (4) news bulletin or local story without affiliation to a hospital, university or private clinic.
Each website was classified by the country of the author.
Content of Information
Claims Regarding Benefits of the Procedure
We searched each website’s content for specific claim regarding the benefits of PRP, which included:
A. Pain improvement
B. Improvement in stiffness
C. Improvement in function
D. Superiority over HA/steroids
E. The ability to delay/eliminate the need for surgery
F. The ability to slow down/limit progression of arthritis
G. The ability to recover cartilage/ prevent cartilage loss
Mention of Associated Risks
The websites were assessed for the inclusion of a description of possible risks or adverse events associated with the injected PRP or the procedure itself.
Description of Technique
The websites were assessed for including a detailed description of the PRP preparation and delivery techniques and the need to draw blood from the patient.
Patient Eligibility Criteria
The websites were assessed for a description of inclusion and exclusion criteria for PRP therapy in general.
References to Peer-Reviewed Literature
To qualify, the websites needed to contain references to peer-reviewed literature regarding PRP and associated advantages and/or disadvantages.
Ability to Make Appointments
Websites were evaluated for offering readers the ability to make an appointment with a physician or other health care provider via online consultation, phone, or email. The information needed to be available on the website itself and not by links to a different website.
Additional Information and Statements
Each website was evaluated for:
A. Mentioning costs and prices for the procedure
B. Acknowledgment of the lack of adequate evidence of the long-term benefits
C. Acknowledgment that PRP is not treatment of choice for end-stage osteoarthritis
D. That the treatment may require several injections to provide a clinical effect, and that the effect may be only temporary
Readability
For each document, the Flesch-Kincaid Grade Level 29 and Flesch Kincaid Reading Ease 30 were used to analyze written material readability in terms of academic grade levels. The Flesch Reading Ease formula describes ease of reading and has been widely used in evaluating medical literature. The Flesch-Kincaid Grade Level formula indicates the required academic grade of the reader and is determined by sentence length and syllable count. We used a free online calculator (http://www.readabilityformulas.com/) to calculate the readability level using the 2 readability methods.
Statistical Analysis
Statistical analysis was carried out using SPSS software V.24 (IBM Corp., Armonk, NY). Comparisons of categorical data according to the search engine (Google, Yahoo!, Bing) were performed using analysis of variance (ANOVA). Comparisons of categorical data from the websites according to authorship (private physician/group vs. other) were performed using the chi-square or Fisher’s exact test when appropriate, depending on the sample size for each comparison.
Results
After screening for duplicates and filtering the results according to the inclusion criteria, 181 sites were left for analysis. A total of 152 were from private physicians or groups and the additional 29 from universities (5), professional society (2), industry/ manufacturer (3), and news bulletin/local story (19). Since the majority of the websites were authored by private physicians or groups and the other authorship groups were significantly smaller, we combined the smaller groups into a single one labeled “Non-private.” Of the 181 results, the leading countries of origin were the United States (146), Canada (10), Australia (10), and the United Kingdom (7). There was no statistically significant difference between the search engines with regard to the type of authorship (P = 0.746, Fisher’s exact test), general characteristics, specific claims, and readability parameters ( Table 1 ).
General Characteristics of the Included Websites, Specific Claims, and Readability Parameters: A Comparison between the 3 Search Engines.
The preparation and delivery techniques were described in similar rates among both private and nonprivate sites, and as was the need for blood drawing prior to the procedure and contraindications (P = 0.369 and P = 0.194, P = 0.848, respectively). Costs were only mentioned by the minority of the websites (11.6%), and mainly by the nonprivate websites (27.6% vs. 8.6%, P = 0.003), yet private websites were more likely to provide contact information (31% vs. 93.4%, for nonprivate and private websites, respectively, P < 0.001).
Most of both private sites and nonprivate sites claimed that PRP injections for osteoarthritis of the knee improves patients’ pain (80.3% vs. 75.9%, P = 0.275), and nearly half stated that the therapy leads to functional improvement (44.8% vs 43.4%, P = 0.952). Overall, 33.6% of private websites and 21.4% of nonprivate websites claimed that PRP treatment may limit or delay the progression of arthritis. A total of 42.8% of the private websites and 27.6% of nonprivate websites have stated that the procedure can delay or eliminate the need for future surgery (P = 0.222). A similar proportion in both groups have reported the risk for adverse effects such as swelling and regional pain (39.5% and 31.0% for private and other websites, respectively, P = 0.391). In all, 29.3% of all websites claimed that PRP therapy is superior to intra-articular steroids or HA.
Both website groups were unlikely to mention that PRP therapy is not the treatment of choice for end stage knee osteoarthritis (7.9% of private groups and 17.2% of the other sites, P = 0.175), or to state that patients with less advanced disease may benefit more from the treatment (11.8% of private groups and 20.6% of the other sites, P = 0.310). The possible requirement for several treatment courses was mentioned mainly by the private websites (47.4% and 27.6%, respectively, P = 0.049). Private websites were less likely to refer to peer-reviewed literature (18.4% vs. 41.4%, P = 0.002) and were more than 3 times less likely to mention lack of adequate evidence in support of PRP therapy (13.2% vs. 48.2%, P = 0.005).
The Flesch-Kincaid Readability Test and the Flesch-Kincaid Grade Level test (FKGL) were calculated to assess readability of the texts in both private and nonprivate websites. Scores were similar for all indexes ( Table 2 ). The American Medical Association’s recommended sixth grade readability guidelines for health information or the mean U.S. adult reading level of eighth grade. 31 The mean FKGL was found to be 7.8 ± 2 for all websites analyzed, reflecting a grade level between seventh and eighth. Of the 181 webpages, 90 (53.3%) had a readability score equal to or above the eighth-grade level, which is the average reading level of U.S. adults. The Flesch-Kincaid Reading Ease mean score was found to be 52.6 ± 14.7, which corresponds to a “fairly difficult” reading ease.
General Characteristics of the Included Websites, Specific Claims, and Readability Parameters: A Comparison Between Private and Nonprivate Websites.
Discussion
Despite a large number of studies examining the efficacy of PRP injections for knee OA, some of which show favorable results, the evidence is of variable quality and concerns include the large heterogeneity of the patients, lack of standardization in study design and outcome measures, variations in treatment techniques, and inconsistencies in the findings of these studies.32,33
The AAOS standpoint on the use of PRP for symptomatic knee osteoarthritis is that some initial research is being done to evaluate the effectiveness of PRP in the treatment of the arthritic knee, and it is still too soon to determine if this form of treatment will be any more effective than current treatment methods.34,35 Therefore, the AAOS cannot recommend for or against the use of PRP and the strength of recommendation was deemed inconclusive. Similarly, the National Institute for Health and Care Excellence (NICE) in the United Kingdom published guidelines 32 in May 2014, which state that the evidence on efficacy is inadequate in quality.
A previous study determined that the quality and validity of information available online regarding osteoarthritis showed a poor overall quality, 36 although the authors did not aim to address different treatment modalities. Taking that into consideration, we hypothesized that the information available online concerning PRP for knee osteoarthritis, will include specific unsubstantiated claims in favor of the treatment, particularly when authored by private physicians or groups.
Our analysis revealed that the majority of the websites in all 3 search engines and in both groups, private and nonprivate, was originated from the United States and the majority belonged to private physician/group consistently.
There were several significant differences between the private and nonprivate sites that may be explained by the different nature of the sites and the tendency toward marketing in the private websites. While the preparation and delivery techniques and specifically the need for blood drawing was described in similar rates among both private and nonprivate sites, costs were only mentioned by few, and mainly by the nonprivate websites: 8 (27.6%) versus 13 (8.6%), P = 0.003. A recent study found by Piuzzi et al. 33 found that the price per a single PRP injection was $714 in average, as currently marketed to patients across centers nationally in the United States. In other words, this is an expensive treatment, especially in light of the fact that most insurance companies do not provide coverage for it. Private websites were less likely to refer to peer reviewed literature (28 [18.4%] vs. 12 [41.4%], P = 0.002) and were more than 3 times less likely to mention lack of adequate evidence (20 [13.2%] vs. 14 [48.2%], P = 0.005). Additionally, 42.8% of the private websites (opposed to 27.6% from the other websites) have stated that the procedure can delay or eliminate the need for future surgery.
A significant proportion of both private sites and nonprivate sites claimed that PRP therapy for osteoarthritis of the knee improves patients’ pain (80% and 75%, respectively) and function (43% and 44%, respectively). However, less than 10% stated that the effect may be only temporary, and less than 50% described a possible need for several injections. A recent systematic review showed no long-term statistically significant improvement in patient validated outcomes and secondary outcomes following PRP administration. 37 The superiority over HA or steroid intra-articular injections was claimed in 28% of the private websites. This proposed claim is questionable, particularly in the long term.8,10
The cost-effectiveness of PRP therapy for knee OA, in terms of postponing or preventing the need for further intervention and particularly surgery, has not been established.33,38 Nevertheless, nearly 43% of the private websites claimed that the treatment can eliminate or delay the need for surgery, 33% claimed that it can slow the progression of the arthritic degeneration, and 40% claimed it can promote cartilage recovery or prevent cartilage loss.
Only approximately 10% of the websites included a description of the indications for injections. The minority of the websites cared to state that the treatment should be directed to patients with mild arthritis (11% of websites), and that it may not be beneficial for more advanced or end-stage arthritis (7.9% of websites). Several studies have clearly shown that PRP is more effective in young patients with early or moderate stages of arthritis but has only limited effect in cases of advanced osteoarthritis15,39 and is clearly not the treatment of choice in those cases.
Limitations of this study relate first to the searches that were conducted. We only used a single search string; other search terms may result in identification of different information, however we assumed this does not detract from the key messages of this study. Second, retrieval of relevant websites may have been limited from a global perspective by restricting our searches to websites in the English language. Additionally, countries with considerably different health care systems were not analyzed separately and that fact can certainly affect the distribution of the different authorship groups. Finally, the internet is dynamic and the information appearing on websites is subjected to constant changes.
The current study demonstrates that claims that have not been appropriately substantiated in the literature are presented online and therefore cannot be presented to the public as absolute advantages of this therapeutic modality. Health-related information needs to be evidence based, precise, and unambiguous in order to provide the patient with the most valid and reliable information, as more credible sources become harder to distinguish from less credible ones.40,41
This is of particular importance when promoting a relatively new treatment modality. It must be presented with caution, striking out the possible benefit but at the same time acknowledging the possible drawbacks and concerns. In addition, information presented online should be uniformly consistent with the existing literature and supported by an adequate amount of relevant references. As more and more data will become available in the near future, PRP may become an important treatment modality for osteoarthritis and the potential benefits proposed today for its use may be further supported with stronger clinical evidence. Until then, these hypothetical benefits cannot be presented to the public as absolute advantages of the treatment. As the interest in PRP continues to increase, health professionals have a responsibility to use the internet as a means of patient education instead of a marketing tool.
Footnotes
Acknowledgments and Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The authors gratefully acknowledge the assistance of Mrs Neta Zalcberg in the preparation of this article.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
