Radiotherapy for prostate cancer: DISCERN quality assessment of patient-oriented websites in 2018

Prostate cancer is the most commonly diagnosed cancer in the male population worldwide [1]. Radiotherapy for localized prostate cancer remains one of the mainstays among treatment approaches. Other fields of application are biochemical recurrences after prostatectomy, postoperative radiotherapy in high-risk situations and palliative settings e.g. painful bone metastases [2, 3]. As most patients do not have personal experiences with radiotherapy, a lack of information on risks and benefits is immanent. A recent Canadian survey revealed that more than 90% of men with newly diagnosed prostate cancer prefer to either play an active or collaborative role in treatment decision-making [4]. Thus, gaining valid medical information before initiating treatment seems essential for these patients to improve their decision-making.

Today, a significant number of patients with prostate cancer [5, 6] and radiation oncology patients [7, 8] utilize the Internet to obtain medical information. For cancer patients in general, Castleton et al. reported Internet utilization in approximately two thirds of patients [9]. However, the quality of online information for patients with prostate cancer is variable [6, 10‐12]. For instance, Ilic et al. found the quality of websites on prostate cancer screening to be highly variable but mostly of poor standard [10]. Similarly, others found shortcomings in quality and accuracy of Internet health information for the special topic of prostate cancer and proton therapy [11]. Noteworthy, Shaverdian et al. showed that 39% of patients who selected the Internet as their primary information source reported their actual treatment experience to be worse than expected. Patients who had initially gathered medical information by consulting their urologists, radiation oncologist or even other patients reported a significantly better treatment experience [6]. On the other hand, a recently published analysis revealed a high rating in quality, accessibility and usability of websites on prostate cancer in general [12].

To our knowledge, there is no current analysis of consumer health information on the Internet for patients with prostate cancer in need of a radiation therapy. In order to close that gap, our goal was to evaluate websites with the focus on prostate cancer and radiotherapy.

First, we analyzed the consistency of our data (i.e. exclusion rates, duplicate websites and temporal changes in search rank). Web search results obtained via Google seemed adequately directed at laypersons (one website excluded) and were highly consistent, as no duplicate results occurred and only a single new webpage was evident within the 20 hits in our second search after three weeks (Fig. 1A). In contrast, exclusion rates (n = 8 and 25) and the incidence of double hits (n = 8 and 10) and of novel search results/ranks (n = 10 and 14, Fig. 1B, C) were higher when Bing and Yahoo were employed, respectively.

Overall, the quality of information on websites about prostate cancer and radiotherapy was good, with a mean DISCERN Plus score of 55.1 ± 10.0, 51.8 ± 10.3 and 50.7 ± 10.1 for Google, Yahoo and Bing, respectively. The scores ranged from a minimum of 34 to a maximum of 74. 13% of all websites were rated as excellent, 31, 48 and 8% as good, fair and poor, respectively. No website was rated as very poor. Figure 2 shows the item-based results of the DISCERN score for the first search results on August 22nd 2017 of all three search engines (n = 39 websites in total). The items with the lowest scores were related to website content concerning sources/references, up-to-dateness and quality of life issues (question 4–6 and 13, Table 1). The overall quality of websites (DISCERN Plus score) was neither dependent on the choice of search engines nor did we observe a significant temporal change in quality (Fig. 3). The main categories of websites retrieved from the first search were charity/NGO sites (46%), followed by sponsored medical news sites (28%), hospital/university sites (20%) and governmental sites (6%). Websites operated by charity organizations had significantly higher DISCERN Plus scores (mean score: 55.5 ± 9.3) compared to hospital sites (mean score: 47.3 ± 9.6, p < 0.042) and medical news sites (mean score: 46.1 ± 6.1, p < 0.009), respectively. Most websites originated in the USA (46%) followed by the UK (33%), Australia/New Zealand (15%) and Canada (6%). The JAMA benchmark criteria were fulfilled for all four sections in 13%, for three, two and one section(s) in 13, 31 and 40%, respectively. One website did not fulfill any of the JAMA criteria. Only 13% of all websites were HON code certified. The median ALEXA global traffic rank was 62,375 with a minimum of five and a maximum of 4,710,605.

Of all analyzed websites, the vast majority had a clear focus on radiotherapy for prostate cancer; four websites provided only information for radiotherapy in general. All analyzed websites had a focus on curative teletherapy (i.e. definitive radiotherapy to the prostate or to prostatic bed in a postoperative or salvage setting), 76, 51 and 22% of all websites mentioned brachytherapy, active surveillance and palliative radiotherapy (i.e. radiotherapy to other sites than the prostate or prostatic bed, e.g. bone metastases), respectively. In 57% the procedure of radiotherapy starting from first preparation, daily treatment and aftercare was described in detail. Special radiation techniques like “hypo-fractionation”, “Intensity modulated Radiotherapy (IMRT)”, “Image guided Radiotherapy (IGRT)” and “proton therapy” were mentioned in 37, 72, 27 and 31% of all analyzed websites, respectively.

In recent years, the Internet has gained importance as a source for health-related information. Several study groups confirmed this fact in patients with oncological diseases in general [9, 23], as well as for patients with gynecologic cancers [24], breast cancer [19, 20], colorectal cancer [25], bladder cancer [26], laryngeal cancer [27] and prostate cancer [12] in particular. The increasing amount of data published within the last two years highlights its clinical relevance.

For prostate cancer in particular there is a high need for information because several treatment strategies exist. Radiotherapy represents a major component for the management of this oncologic entity. Thus, we performed an evaluation of websites addressing information on prostate cancer and radiotherapy. The first 20 websites were subjected to further evaluation as it is known that patients rarely browse through further sites when searching the web [13]. Ngyuen et al. showed that 72% of the patients even restricted their search to 1–5 websites only [14]. In a study of Borgmann et al. the list of evaluated websites was limited to 10 sites [12].

Overall, we found a good quality of information on analyzed websites according to the validated DISCERN Plus score (for Google search: mean DISCERN Plus score: 55.1 ± 10.0). Our results are in concordance with findings from studies analyzing website quality of other cancer entities, returning a mean DISCERN Plus score of 50 [27], 52 [25], 42 [28] and 42 [20], respectively. Notwithstanding, the only other comparable study evaluating websites with general information on prostate cancer, obtained higher DISCERN Plus scores (mean 77) [12]. Taken together, websites with information on prostate cancer seem to be of slightly better quality compared to websites covering other cancer entities. A potential reason why the web search by Borgmann et al. returned webpages of higher quality might be associated with the use of a more general and popular search term. This is reflected by a lower ALEXA traffic rank for prostate cancer sites in general (median ALEXA traffic rank in the study of Borgmann et al.: 2718) compared to our focus on radiotherapy and prostate cancer (median ALEXA traffic rank: 62375). In our study, webpage quality (DISCERN Plus score 2.14 ± 1.03) of less frequently visited websites (n = 14, ALEXA traffic rank < 62,375) was not significantly different from more frequently visited sites (n = 15, ALEXA traffic rank > 62,375, DISCERN Plus score 1.67 ± 0.98). We did not find an association of ALEXA traffic rank and superior DISCERN score (data not shown).

In our analysis only 13% of websites were HON code certified. Most other studies evaluating websites on cancer entities in general found higher certification rates ranging from 20 to 38% [12, 19, 20, 26, 29‐31]. The vast majority of these studies showed that a higher HON code certification was not necessarily accompanied by a better content quality, as determined by the DISCERN Plus score. Borgmann et al. presumed a lower HON code certification frequency for websites on more specific topics, e.g. prostate cancer and radical prostatectomy. This phenomenon could also account for the low number of HON code certified websites in our study, which is in concordance with findings by the study group of Alkhateeb indicating 17% of HON code certification in their focused web search on prostate cancer and surgery [32].

The websites operated by charity organizations were of superior quality concerning the DISCERN score compared to hospital sites and sponsored medical news sites. In line with this finding, Liebl et al. also found websites of non-profit providers and self-help groups of superior quality compared with profit-driven websites for cancer patients in general [33]. Moreover, an earlier analysis of our study group also revealed better results for hospital and NGO websites compared to sponsored medical news sites for webpages with information on breast cancer [20]. Thus, currently non-commercial sites should be preferentially recommended to patients seeking information on prostate cancer and radiotherapy.

No significant differences in the DISCERN Plus score, JAMA benchmark criteria and HON code certification were found between the three search engines (data not shown). Search results obtained via Google were greatly directed at laypersons and showed a high grade of consistency (i.e. exclusion rates, duplicate websites and temporal changes in search rank), in contrast to the other search engines. These properties might render Google a reliable choice for laypersons searching for information on prostate cancer and radiotherapy.

Interestingly, a large percentage of analyzed websites provided information on modern radiotherapy techniques like IMRT (72%). Also, websites mentioned brachytherapy in the setting of curative treatment in 76% and the possibility of no treatment (active surveillance) in 51%. Despite the well-balanced information on modern treatment options for localized disease, we found a significant lack of information on palliative radiotherapy, which was only mentioned by 22% of websites.

Websites on radiotherapy and prostate cancer directed at laypersons have the potential to sufficiently inform patients about general treatment options. The fact that we were unable to find a simple strategy for the identification of high-quality websites (i.e. HON code certification, JAMA benchmark criteria, ALEXA ranking, different search engines or country of origin and DISCERN Plus score) emphasizes the responsibility of the treating physicians to interpret and rank the vast quantity of information.