- Email: [email protected]
Prostate Cancer on the Internet—Information or Misinformation?
Prostate Cancer on the Internet—Information or Misinformation? Peter C. Black* and David F. Penson† From the Departments of Urology, University of Washington, Seattle, Washington (DFP), and University of Southern California/Norris Cancer Center, Los Angeles, California (PCB)
Purpose: We assessed the quality of information available to patients on prostate cancer on the Internet. Materials and Methods: The search engine Webcrawler® was used with the search term “prostate cancer” to generate a list of 75 websites which were reviewed for currency, disclosure, attribution, interactivity and content. A rating tool was designed including 50 elements considered essential for a comprehensive review of prostate cancer, and each website was judged for degree of coverage and accuracy (each rated on a scale of 1 to 3) of information for each element. Results: Of the 75 sites 39 contained information about prostate cancer. Only 9 sites indicated a date of last update within 6 months. References were rarely given (in 5) and a disclaimer was provided on less than half of the sites (18). The sites covered a mean of 24 elements (range 6 to 43) with a mean coverage rating of 1.0 to 2.6 (1.8 overall). Of 943 elements covered on 39 sites, 94% were completely correct, 5% were mostly correct and 1% was mostly incorrect. Conclusions: The information on the Internet is of sufficient quality to aid in patient decision making. However, there are numerous shortcomings especially related to currency, disclosure and attribution. Degree of coverage is highly variable and there is a deficiency in balance of evidence found on many sites. The urologist needs to be aware of such shortcomings when counseling patients on prostate cancer. Key Words: prostatic neoplasms, internet
he quantity of information on the Internet related to prostate cancer is increasing rapidly, but the quality of this information has not been evaluated systematically.1 Concurrently, patients are becoming more facile with the Internet and are turning to it often as an important source of medical information. Up to 45% of patients with prostate cancer are using the Internet for self-education.1– 4 In some instances the information found on the Internet appears to be a key determinant in medical decision making.3,4 Information gathered from the Internet is usually not peer reviewed and can be posted without proper attribution.1 Many patients are probably not aware of these concerns.1,5 This is particularly problematic in prostate cancer, as it is a common malignancy associated with many controversial issues. The newly diagnosed patient is faced with numerous treatment options backed by minimal clinical evidence, further intensifying the need for useful information.6,7 In this study we assessed the quality of information available on prostate cancer on the Internet through a structured review of numerous websites. Specifically we wanted to determine if the quality of this information was adequate to support patient decision making.
T
Submitted for publication June 11, 2005. * Correspondence: Department of Urology, University of Washington, Box 356510, 1959 NE Pacific St., Seattle, Washington 98195 (telephone: 206-543-3640; FAX: 206-543-3272; e-mail: pblack@u. washington.edu). † Financial interest and/or other relationship with Sanofi-Aventis, Icos/Lilly, Amgen, Dendreon, Pfizer and Boehringer-Ingelheim.
See Editorial on page 1604.
0022-5347/06/1755-1836/0 THE JOURNAL OF UROLOGY® Copyright © 2006 by AMERICAN UROLOGICAL ASSOCIATION
MATERIALS AND METHODS An Internet search using the term “prostate cancer” without modifications or advanced search techniques was performed with the search engine Webcrawler®.3 Webcrawler® was chosen based on its search technology, which includes an evaluation of search results from other search engines. It reviews the information content of Web pages, their metatags, anchor text and link popularity to generate a search result that emphasizes relevance. The search yielded 390,770 websites, of which the top 75 sites were included in analysis. These websites were evaluated using a structured rating tool developed according to the “Criteria for Assessing the Quality of Health Information on the Internet” (1997)8 and standards and methodologies proposed by Bertrand et al for the systematic review of the quality of medical websites.9 Each website was evaluated by the research team for 5 central components, namely currency, disclosure, attribution, interactivity and content. The parameters describing the first 4 components are outlined in Appendix 1. Balance of evidence (Appendix 2) was evaluated as a component of attribution based on 4 controversial issues specifically related to prostate cancer: prostate cancer screening, watchful waiting vs immediate primary treatment, radiation treatment vs surgery and immediate vs delayed hormonal treatment. Between 3 and 5 elements of information were listed for each issue, and the websites were reviewed to evaluate the degree to which each statement was covered. The coverage scale outlined in table 1 was used. Content was measured with a rating tool designed to include 50 elements of information considered essential for a comprehensive review of prostate cancer. These elements
1836
Vol. 175, 1836-1842, May 2006 Printed in U.S.A. DOI:10.1016/S0022-5347(05)00996-1
PROSTATE CANCER INFORMATION ON INTERNET TABLE 1. Rating scales used to evaluate coverage and accuracy Description Coverage: 1 2 3 Accuracy: 1 2 3
Minimally addressed More than minimally addressed Mostly addressed Mostly incorrect Mostly correct Completely correct
1837
TABLE 2. Sites for which individual aspects of disclosure are present or absent
Name of organization Nature of organization Authorship Purpose of website Source of support
No. Indicated
No. Not Indicated
38 36 22 38 33
1 3 17 1 6
RESULTS
choosing radiation or surgery as a treatment modality for prostate cancer, but only 12 discussed at least 3 of the 5 rating elements more than minimally or completely. Five sites discussed the issue of immediate vs delayed hormone ablation, but only 2 in a balanced fashion. References were listed in a complete fashion on 5 sites, incompletely on 6 and not at all on 28. A disclaimer was provided on 18 of 39 sites. A contact or feedback mechanism was made available on 38 sites. The sites covered a mean of 24 elements (range 6 to 43) with 6 sites covering more than 35 elements (fig. 1). The mean coverage rating of individual sites ranged from 1.0 to 2.6 with an overall coverage rating of 1.8 (fig. 2). In total 943 elements were covered on the 39 sites, and of these 25% were “mostly addressed,” 34% were “more than minimally addressed” and 40% were “minimally addressed.” The mean total coverage score was 45 out of a possible 150. The scores ranged from 6 to 88 with 9 sites scoring 1 to 25, 14 sites scoring 26 to 50, 13 sites scoring 51 to 75 and 3 sites scoring greater than 75. Figure 3 shows how the scores correlated to the type of site. Each element was covered on a mean of 19 sites (range 4 to 34). Categorizing the elements revealed that the elements of the given categories were covered on a mean of 18.8 to 24.1 sites for all categories except hormone refractory disease, for which the elements were covered at least minimally on a mean of 8.7 sites (fig. 4). Comparing coverage scores for elements categorized by modality of treatment (radical prostatectomy, external beam radiation, brachytherapy and endocrine treatment), a statement can be made about the balance with which these categories are covered on any given site. A total of 33 sites
The 75 sites included 59 original sites and 16 dead links. Only 39 of the 59 sites had information about prostate cancer, and these 39 sites were evaluated in detail. Six sites were managed by academic institutions or government organizations, 11 were nonprofit organizations, 12 were commercial, 4 were sponsored by individual physicians or private health care organizations, 4 were personal sites including self-help groups, and 2 were news sites. A date of last update was not indicated in 24 of 39 sites. One site had been updated within 2 weeks, 4 within 1 month and 4 within 3 months. Six sites had been updated more than 6 months before the date of review. The data on disclosure are summarized in table 2. Balance of evidence was evaluated using the rating tool designed for this purpose. A total of 25 sites discussed prostate cancer screening, but only 10 described at least 2 of the 4 elements more than minimally or completely. There were 24 sites that discussed watchful waiting and 22 discussed at least 2 of the 3 rating elements more than minimally or completely. The 25 sites addressed the decision process for
FIG. 1. Number of sites that cover specified range of elements. Mean number of elements per site is 24, with individual elements covered on range of 6 to 43 sites.
were derived from the corresponding chapters in Campbell’s Urology10 and are listed in Appendix 3. A minimum of 5 elements had to be covered for a site to be included in the survey. The categories of elements included epidemiology/ etiology/natural history, diagnosis/staging/pathology, radical prostatectomy, external beam radiation, brachytherapy, endocrine treatment, hormone refractory prostate cancer and alternate treatments. Each website was judged for degree of coverage and accuracy of information for each element found on the website using the scales described in table 1. The raw data for coverage generated with this rating tool were analyzed with 3 methods. The number of elements covered by each website was tabulated. Then the mean rating on the scale of 1 to 3 for all elements covered on a site was calculated. Finally, a score was calculated for each website that consisted of the sum of the ratings for each individual element covered. The maximum possible score for 50 elements rated on a scale of 1 to 3 is 150. The scores obtained for each category of elements within a single website could be compared by adding together the ratings within a category to get a sum for that category. The sums could not be compared directly between categories because each category had a different number of elements. Therefore, each score was expressed as a percentage (percentage score) of the total possible score for that category, and the percentages were compared between categories, providing further information on the balance of evidence given on that website.
1838
PROSTATE CANCER INFORMATION ON INTERNET
FIG. 2. Number of sites with mean degrees of coverage falling in specified limits. Overall mean was 1.8, with individual sites ranging from 1.0 to 2.6.
described at least 1 of the 4 standard treatment modalities with at least 20% of the maximum coverage score for that category. Of these 33 sites 1 site described only 1 of the 4 treatment modalities (brachytherapy), while 7 sites left out at least 1 treatment modality completely (1 radical prostatectomy, 2 brachytherapy, 5 endocrine treatment and 1 site left out 2 modalities). If only sites that discussed all 4 treatment modalities are included (33), there were discrepancies in the amount of information provided on each modality. An imbalance of evidence, as indicated by a 20% difference (more or less) in the percentage scores between 1 category and the next closest category, was present in 13 of 33 sites. This consisted of overrepresentation or underrepresentation of radical prostatectomy on 2, external beam radiation on 2, brachytherapy on 3 and endocrine treatment on 5 sites. Of 943 elements covered in 39 sites, 94% were “completely correct,” 5% were “mostly correct” and 1% was “mostly incorrect.” The mean accuracy rating achieved was 3 in 17 sites, 2.9 in 15 sites, 2.8 in 5 sites and 2.7 or 2.6 in 1 site each. The overall accuracy rating was 2.9. The most common
FIG. 3. Median coverage score for each type of website. Mean score for all sites was 45, range 6 to 88. univ/gov, university and government. comm, commercial.
FIG. 4. Mean number of sites covering elements of each category. Number ranges from 8.7 for hormone refractory disease (HR) to 24.1 for diagnosis and staging (diag/stag). epid/etiol, epidemiology and etiology. EBXRT, external beam radiation. brachy, brachytherapy. endo, endocrine treatment.
points of error were the definition of Gleason sum (defined completely in 27 sites, with 20 “completely correct” and 7 “mostly correct”) and the guidelines for PSA screening (defined completely in 16 sites, with 12 “completely correct” and 4 “mostly correct”). Otherwise all ratings for “mostly incorrect” and “mostly correct” were scattered evenly among the 50 elements. Accuracy did not differ by type of site. DISCUSSION This evaluation of the quality of information on prostate cancer on the Internet reveals a variety of important findings related to currency, disclosure, attribution, interactivity and content.11–13 Approximately two-thirds of sites did not indicate the currency of the information presented on the site. This would appear to be a fundamental oversight in any biomedical medium, as the information content changes rapidly. Only 9 sites had been updated within 6 months, which would be an appropriate interval for this kind of information.11 Disclosure is important on web pages because it allows the reader to judge the intent and objectivity of the information, as well as its source. Although the name and nature of the organization constructing the website, and the purpose of the website and the source of sponsorship were clearly indicated in most sites, actual authorship was listed on just more than half of sites. Attribution allows the reader to assess the quality of the information being provided. While a list of references is never absent from a peer reviewed publication, only 5 sites included a complete list of references, and 28 had no references. Although most patients may not pursue further references, it is essential to give the reader the opportunity to verify independently the information offered.14 Similarly, a description of the validity of information (level of evidence) was addressed on only a small number of sites (7 of 39), and in none with any sort of completeness. Every site that could potentially be used for patient education should be expected to include a disclaimer that a patient needs to discuss any newly found information with a health care provider to
PROSTATE CANCER INFORMATION ON INTERNET apply it to his or her individual case, yet only half of the sites in the current study did this. Balance of evidence is a more difficult parameter to assess, but it may be the most important qualitative parameter after content. The results of the current study indicate that there is an adequate balance across all sites (as measured by the number of sites covering elements within the different categories), except information on hormone refractory disease, but that there is a significant imbalance of evidence within individual websites. The coverage of the 4 primary treatment modalities (radical prostatectomy, external beam radiation, brachytherapy and endocrine treatment) was poorly balanced on approximately a third of sites. Endocrine treatment was overrepresented or underrepresented more often than the other modalities, but for the most part the oversights were equally distributed among the 4 modalities. The imbalance was further evident in the coverage of 4 controversial yet central questions on prostate cancer diagnosis and management. One issue was presented in a balanced fashion on only 2 sites, 2 others on a third of sites and the fourth on two-thirds of sites. The issue of immediate vs delayed hormone ablation proved to be a poor example for measuring balance of evidence, since only a small number of sites addressed this topic. However, screening for prostate cancer, the advantages and disadvantages of watchful waiting, and a comparative evaluation of surgery vs radiation treatment are key issues that are not covered in a balanced fashion. The ultimate factor in assessing a website’s quality is its content, which can be looked at as coverage and accuracy. If the content is deficient, the other parameters will not make it a useful site. Hellawell et al have also looked specifically at the information provided on prostate cancer on the Internet and attempted to quantify it. They scored websites on a scale from 10 to 100 and found a mean score of 50.7.1 In the current study while some sites provide excellent coverage of many facets of the disease, others are limited in scope. Six sites were clearly superior to all others with respect to the amount of coverage (more than 35 of the 50 elements). Although the risk of such exhaustive coverage is to overwhelm the reader, in general these sites are carefully constructed and divided into sections that allow easy browsing through relevant information and omission of topics less pertinent to the reader. Interestingly, the top 6 sites varied from public service sites such as the Australian Lions Society and selfhelp groups like UsTOO, to government institutions such as the National Cancer Institute and industry sponsored sites such as Prostatecancer.com (AstraZeneca). However, the percentage scores, another measure of coverage, were higher for the sites sponsored by universities/government and nonprofit organizations compared to those sponsored by private physicians/hospitals and commercial sites. The 8 sites that covered less than 15 elements were mostly sites promoting a specific aspect of prostate cancer or promoting a product. Accuracy of information was excellent across most websites with 99% of information being “mostly” or “completely correct.” Only 6 sites had an accuracy rating of less than 2.9 (best possible 3.0). Although these data on accuracy are for the most part reassuring, the goal should be to avoid any report of false information. In addition, high accuracy is tempered by poor balance, which may be considered equivalent to error of omission15 and can generate a misleading impression.
1839
The main limitation of the current study is the lack of validity of the survey tool and the subjectivity of content elements. We have chosen one search engine from many available, and we have evaluated a relatively small number of sites out of the overall number of sites available. Since these include the most frequently visited sites, we believe that they are representative of what patients are finding. In addition, there are aspects of website design such as readability and user-friendliness that we have not attempted to define, but these also will have a significant impact on the usefulness of a website to an individual patient. We have made no attempt to define which patients are using the Internet and to what purpose and, therefore, are not able to predict what information should be offered and how it should be best presented. Other authors have addressed the question of what could be done to improve the quality of information on the Internet. Pautler et al suggest that national organizations such as the American Urological Association develop a list of approved/endorsed websites, that guidelines be developed for the construction of balanced and responsible websites, and that physicians learn to guide their patients to appropriate websites.3 The American Medical Association offers guidelines for website design16 and a system of website recognition has been developed by organizations such as HONcode.17 Although the quality of Internet information is important, physician familiarity with these data is equally important. Physicians continue to have the greatest influence on patient decision making,7 and it would become us to use the Internet as a counseling tool for suitable patients. CONCLUSIONS The information on the Internet is of sufficient quality to aid in patient decision making. However, there are numerous shortcomings especially related to currency, disclosure and attribution. Degree of coverage is highly variable but the covered content is accurate. The most significant shortcoming is the deficient balance of evidence found on many sites. The urologist needs to be aware of such shortcomings when counseling patients on prostate cancer, and should be able to guide the patients to up-to-date websites that offer balanced information from identifiable, reputable sources. APPENDIX 1 Summary of parameters used to rate individual websites Parameter Currency/date of last update Disclosure: Name of organization Nature of organization Authorship Purpose of website Source of support Attribution: Balance of evidence List of references Validity of information Disclaimer Interactivity: Contact/feedback mechanism Chat room/bulletin boards
Rating Less than 2 wks/less than 1 mo/less than 3 mos/less than 6 mos/greater than 6 mos/not indicated Present/absent Present/absent Present/absent Present/absent Present/absent See Appendix 1 None/incomplete/complete None/incomplete/complete Present/absent None/e-mail/telephone/FAX/ mailing address Present/absent
1840
PROSTATE CANCER INFORMATION ON INTERNET
APPENDIX 2 Rating Tool for Balance of Evidence BALANCE OF EVIDENCE 1. Screening for prostate cancer ● annual DRE and PSA recommended for males over 50 years old; for black males and other high risk males over 45 years old ● no randomized studies have been completed to show an advantage in overall survival by screening with PSA and DRE ● morbidity and cost of tests, including PSA and prostate needle biopsies ● there is no evidence that early aggressive treatment reduces mortality 2. “Watch and wait” versus immediate treatment ● if low grade (Gleason score), low volume disease, low clinical stage ● if patient older, more comorbidities ● then may not need treatment; more likely to die from other cause, may suffer adverse effects of treatment without experiencing benefits of treatment 3. Radiation vs brachytherapy vs radical prostatectomy for organ confined disease ● difficult to compare outcomes of radiation and surgery, because radiation cases tend to be a selection of cases that are not surgical candidates, and there is no pathological staging (pT2 vs pT3) available in radiation cases ● there are no good randomized controlled prospective data comparing different forms of treatment; retrospective studies indicate similar local control between brachytherapy and radical prostatectomy in low stage disease after 5 and 10 years; poorer local control after EBRT ● longer followup pending for radiation/brachytherapy ● each has advantages and disadvantages regarding tolerability and adverse effects (list 2 major adverse effects for each treatment) ● “best” treatment needs to be decided in each patient individually 4. Immediate vs delayed hormone treatment ● only 1 study clearly shows benefit to early hormone treatment; this is in patients with positive lymph nodes found during radical prostatectomy ● otherwise no study shows definite benefit with early vs late hormones (ie decreased complication rate or improved survival) ● hormone treatment involves significant adverse effects and possible decrease in quality of life
APPENDIX 3 Rating Tool for Content
EPIDEMIOLOGY/ETIOLOGY 1999: approximately 179,500 new cases, 37,000 deaths; second leading cause of cancer death in men (higher number earlier in the 90s eg greater than 300,000 cases, greater than 40,000 deaths in 1996) demographics: 95% between age 45 and 89, median age 72 highest incidence in blacks, intermediate in whites, lowest in Chinese/Japanese family history: higher risk if higher number of affected relatives and if affected relatives have earlier age of onset some (eg American Urological Association and American Cancer Society) recommend annual DRE and PSA starting at age 50, others (eg Canadian Task Force on the Periodic Health Examination and the United States Preventive Service Task Force) do not initial studies indicate that selenium and vitamin E reduce the incidence of prostate cancer compared to a placebo; a large national trial is currently underway to verify this PATHOLOGY high grade PIN on needle biopsy approximately 30% to 50% risk of finding carcinoma on subsequent biopsies; similar with atypical small acinar proliferation; repeat biopsy should be performed for both entities definition of Gleason grading: based on glandular pattern of the tumor as identified on relatively low magnification; the primary and secondary architectural patterns are identified and assigned a grade 1 to 5 (5 is least differentiated); sum of both is Gleason score, this is an indicator of how aggressive the tumor is DIAGNOSIS/STAGING usually asymptomatic; sometimes associated with voiding symptoms in locally advanced disease and bony pain in metastatic disease, also acute spinal cord compression; some cancers are diagnosed incidentally at the time of TURP screening: most effective method is PSA plus DRE; 18% will be found by DRE, 45% by PSA and 37% by PSA and DRE together diagnostic tests: TRUS: classically shows hypoechoic lesion in periphery, but carcinoma may be hypo-, iso- or hyperechoic; main use of TRUS is for biopsy guidance needle biopsy: classically sextant biopsy: base, mid and apex bilaterally; newer evidence indicates that 8 to 12 may be better Staging Systems Whitmore-Jewett staging system A1 focal/A2 diffuse CaP on TURP/simple prostatectomy B1 palpable but organ confined unilaterally/B2 bilaterally C1 minimal capsular invasion/C2 extensive capsular invasion with hydro or bladder outlet obstruction D1 lymph node mets/D2 distant mets/D3 hormone refractory Clinical TNM staging system T1 nonpalpable ⫽ normal DRE T1a 5% or less/T1b greater than 5% of specimen on TURP/simple prostatectomy T1c by PSA only T2 palpable, organ confined T2a unilateral less than half/T2b unilateral greater than half/T2c bilateral T3 beyond capsule T3a extracapsular CaP/T3b seminal vesicle invasion T4 invasion of bladder neck, external sphincter, levator, rectum, pelvic wall N0 no nodes/N1 regional nodes M1a nonregional nodal mets/M1b bone/M1c visceral mets Staging: pelvic imaging with computerized tomography or magnetic resonance imaging for detection of local extension or presence of lymph node metastases is not routinely useful because of low sensitivity; possible exception in men with locally advanced disease by DRE, PSA greater than 20 or Gleason 8 to 10 bone scan indicated for patients with PSA greater than 20; bone metastases are typically osteoblastic (appendix 3 continued)
PROSTATE CANCER INFORMATION ON INTERNET
1841
APPENDIX 3 continued
NATURAL HISTORY prospective randomized controlled trials investigating the effect of early detection and treatment of prostate cancer on survival have not yet been completed Survival: approximately 20% of men with CaP will die of disease median survival of men who have: - PSA recurrence after local treatment approximately greater than 10 years - metastatic disease approximately 2–3 years - hormone refractory disease approximately 12–18 months watch and wait: some prostate cancers are clinically insignificant and do not require treatment; these can be watched with regular H&P and PSA surveillance; the patient is followed until either symptoms or an increase in PSA leads to treatment RADICAL PROSTATECTOMY (RRP) usual indication: life expectancy greater than 10 years, clinical T1–2 disease; high PSA (eg greater than 20) is not a contraindication by itself as long as no other evidence of metastatic disease PSA is used to follow after RRP - should be undetectable; probability of PSA recurrence correlates to preoperative PSA, Gleason score (in needle biopsy and in final surgical specimen) and pathological stage 5 year progression-free survival rate for T1–2, Nx disease is 69% to 83%; 10 year progression-free survival rate 47% to 78%; greater than 90% progression-free survival for organ confined disease invasion of seminal vesicles and involvement of pelvic lymph nodes are poor prognostic indicators; significance of positive margins is uncertain - only 40% to 50% of patients with positive margins (and no other adjuvant treatment) will have PSA recurrence within 5 years complications: most common early complication is hemorrhage; average blood loss 800 to 1,500 ml operative mortality approximately 0.5% (death within 30 days of RRP) also: rectal injury in less than 1%, wound infection/dehiscence in 1.5%, lymphocele in 3% to 4%, anastomotic stricture in 0.5% to 9.0%, DVT in 1% to 2%, PE in 1% to 2% rate of postoperative incontinence depends on source of data: patient surveys give rate of 19% to 31%, centers of excellence report 5% to 10%; details of surgical technique are thought to be important in determining rate of continence; can improve up to 24 months postoperatively the cavernosal nerves, which are essential for erections, are resected during a standard RRP; in a nerve sparing prostatectomy they can be saved on 1 or both sides of the prostate if this is thought not to jeopardize tumor control return of erectile function depends on potency status prior to RRP, stage of disease, unilateral vs bilateral nerve sparing and age of patient; up to 75% to 80% of previously potent men younger than 60 years with bilateral nerve sparing will retain erectile function; population surveys give lower numbers than are reported from centers of excellence; can improve up to 24 months postoperatively RADIATION-EXTERNAL BEAM alternative to prostatectomy and brachytherapy in T1–2 disease; it has not been definitively shown that survival after external beam radiation is better or worse than prostatectomy or brachytherapy treatment of choice, combined with hormone ablation (for cytoreduction), for T3 disease, but poor progression-free survival rates at 5 and 10 years (15% to 30%) external beam radiation can be offered adjuvantly to radical prostatectomy in three settings: 1. occult residual tumor in pT3 disease, 2. PSA recurrence after RRP with extraprostatic extension, and 3. clinical local failure; should wait 5 to 8 weeks to allow for return of continence and complete healing from RRP most centers use “conventional simulation”; however, newer “3-dimensional conformal delivery” reduces complications (computer mapping of tumor within prostate, patient fitted with plastic mold resembling a cast to limit motion, improve accuracy of aim) short-term adverse effects include diarrhea, rectal irritation, dysuria, frequency chronic complications in 5%: rectal/bladder injury (including rectal bleeding and hematuria), enteritis, urethral stricture, 0.9% incontinence, 40% to 70% impotence RADIATION-BRACHYTHERAPY TRUS guided placement of radioactive I125 or Pd103 into prostate through the perineum; seeds emit radioactivity for 6 (Pd103) to 12 (I125) months; conceptual advantage: higher radiation dose to prostate with less exposure to surrounding tissue; dose greater than 100 Gy can be offered to any patient with clinically localized prostate cancer as an alternative to prostatectomy or external beam radiation; relative contraindications include large prostate, severe obstructive voiding symptoms and s/p TURP combined with EBRT in patients with high risk for extracapsular disease (T2b–T3a, Gleason 7 to 10, PSA greater than 10) short-term results (PSA-free survival) similar to radical prostatectomy, and better than EBRT for organ confined disease; long-term results pending short-term adverse effects include acute urethral irritiation, urinary obstruction (including retention) and proctitis long-term adverse effects include urinary incontinence (higher after TURP), diarrhea and erectile dysfunction (especially older patients); these problems may increase with time popular perception is that brachytherapy causes less impairment of quality of life, but randomized studies are pending; appears to cause less incontinence and impotence but more irritative symptoms ENDOCRINE TREATMENT includes orchiectomy, LH-RH agonists (mainstay), antiandrogens and estrogens; no regimen is superior to any other in survival indication: locally advanced or metastatic disease; also for patients with high grade, organ confined disease who are not candidates for other treatments and patients with PSA recurrence after local treatment androgens are made in the testes and adrenal glands (5% to 10%); only testicular production ceases with LH-RH agonist or orchiectomy; total androgen blockade is LH-RH agonsist or surgical castration plus antiandrogen concurrently; the antiandrogen blocks the effect of testosterone made extratesticularly; there is no evidence that adding an antiandrogen improves time to progression or survival side effects include loss of libido and potency (but not in all men); also hot flashes, osteoporosis, fatigue, loss of muscle mass, anemia, weight gain, elevated liver enzymes; libido and potency maintained in pure antiandrogen therapy intermittent endocrine treatment: theoretical advantage of prolonging survival by delaying progression to androgen independence – subject of ongoing randomized studies initial studies indicate that there is an improvement in quality of life, including sexual function, during the off-therapy interval a period of 6 to 9 months on therapy is usually recommended; the mean off-therapy interval approaches 50% of the duration of the treatment cycle neoadjuvant endocrine treatment: there appears to be no advantage (biochemical recurrence) to 3 months of neoadjuvant hormone ablation; studies are ongoing looking at 8 months of treatment; the rate of positive surgical margins is lower after neoadjuvant endocrine treatment, but the clinical significance is uncertain TREATMENT OF HORMONE REFRACTORY CARCINOMA testosterone level should be checked to evaluate adequacy of hormone ablation - should be castrate levels; hormone ablation is continued during chemotherapy antiandrogen withdrawal can result in temporary clinical response seen as decreased PSA, symptomatic relief; requires serial monitoring of PSA every 4 to 8 weeks; median duration of effect approximately 3 months (appendix 3 continued)
1842
PROSTATE CANCER INFORMATION ON INTERNET
APPENDIX 3 continued
second-line endocrine intervention includes ketoconazole, diethylstilbestrol, corticosteroids or aminoglutethimide cytotoxic agents used include mitoxantrone, prednisone, estramustin, taxol, taxotere, cyclophosphamide, etoposide, vinblastine, doxorubicin first line palliative treatment for symptomatic bone lesions is external beam radiation; strontium 89 can increase symptom-free interval by targeting subclinical disease if neurological deficits, first line treatment is high dose corticosteroids, then external beam radiation; surgery if unstable fracture, progression during radiation or recurrence after radiation future directions of chemotherapy include topoisomerase I inhibitors (TOPO-I), tissue inhibitors of metalloproteases (TIMPS), inhibition of angiogenesis, differentiation agents (eg, Vit D, butyrates), gene therapy and monoclonal antibodies ALTERNATIVE TREATMENTS cryoablation is a rarely performed treatment, accepted currently only for use after local radiation failure; uses probes placed by TRUS guidance; urethral warming catheter to protect from necrosis; advancement of ice ball observed by TRUS with 1 to 2 mm accuracy; extends up to muscularis propria of rectum salvage prostatectomy: used for local recurrence following radiation in select patients; defined as positive biopsy in conjunction with rising PSA and no evidence of metastases; preoperative PSA should be less than 10, (15% vs 86% advanced pathological features)
Each element (first level bullet) was judged for coverage and accuracy according to the scales outlined in table 2.
Abbreviations and Acronyms CaP DRE EBRT H&P LH-RH PSA RRP TRUS TURP
⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽
prostate cancer digital rectal examination external beam radiation therapy history and physical examination luteinizing hormone releasing hormone prostate specific antigen radical retropubic prostatectomy transrectal ultrasound transurethral resection of the prostate
REFERENCES 1. Hellawell, G. O., Turner, K. J., Le Monnier, K. J. and Brewster, S. F.: Urology and the Internet: an evaluation of internet use by urology patients and of information available on urological topics. BJU Int, 86: 191, 2000 2. Diefenbach, M. A., Dorsey, J., Uzzo, R. G., Hanks, G. E., Greenberg, R. E., Horwitz, E. et al: Decision-making strategies for patients with localized prostate cancer. Semin Urol Oncol, 20: 55, 2002 3. Pautler, S. E., Tan, J. K., Dugas, G. R., Pus, N., Ferri, M., Hardie, W. R. et al: Use of the internet for self-education by patients with prostate cancer. Urology, 57: 230, 2001 4. Smith, R. P., Devine, P., Jones, H., DeNittis, A., Whittington, R. and Metz, J. M.: Internet use by patients with prostate cancer undergoing radiotherapy. Urology, 62: 273, 2003 5. Henson, D. E.: Cancer and the Internet. Cancer, 86: 373, 1999 6. Berry, D. L., Ellis, W. J., Woods, N. F., Schwien, C., Mullen, K. H. and Yang, C.: Treatment decision-making by men with localized prostate cancer: the influence of personal factors. Urol Oncol, 21: 93, 2003 7. Patel, H. R. H., Mirsadraee, S. and Emberton, M.: The patient’s dilemma: prostate cancer treatment choices. J Urol, 169: 828, 2003 8. http://www.mitretek.org/hiti/showcase/documents/criteria.html. Accessed April 4, 2002 9. Berland, G. K., Elliott, M. N., Morales, L. S., Algazy, J. I., Kravitz, R. L., Broder, M. S. et al: Health information on the Internet: accessibility, quality, and readability in English and Spanish. JAMA, 285: 2612, 2001 10. Walsh, P. C., Retik, A. B., Vaughan, E. D., Jr. and Wein, A. J.: Campbell’s Urology, 7th ed. Philadelphia: W. B. Saunders Co., 1998
11. Al-Bahrani, A. and Plusa, S.: The quality of patient-orientated internet information on colorectal cancer. Colorectal Dis, 6: 323, 2004 12. Gilliam, A. D., Speake, W. J., Scholefield, J. H. and Beckingham, I. J.: Finding the best from the rest: evaluation of the quality of patient information on the Internet. Ann R Coll Surg Engl, 85: 44, 2003 13. Lee, C. T., Smith, C. A., Hall, J. M., Waters, W. B. and Biermann, J. S.: Bladder cancer facts: accuracy of information on the Internet. J Urol, 170: 1756, 2003 14. Darmoni, S. J., Amsallem, E., Haugh, M., Lukacs, B., Leroux, V., Thirion, B. et al: Level of evidence as a future gold standard for the content quality of health resources on the internet. Methods Inf Med, 42: 220, 2003 15. Fagerlin, A., Rovner, D., Stableford, S., Jentoft, C., Wei, J. T. and Holmes-Rovner, M.: Patient education materials about the treatment of early-stage prostate cancer: a critical review. Ann Intern Med, 140: 721, 2004 16. http://pubs.ama-assn.org/how_principle.htm. Accessed July 8, 2004 17. http://www.hon.org. Accessed March 25, 2003
EDITORIAL COMMENT The authors present a valuable study on content accuracy of prostate cancer websites using Campbell’s Urology as a gold standard. However, there is a question of whether the average male searching for prostate cancer information would retrieve the same 39 sites reviewed here out of a possible 390,000. Moreover, information needs to be delivered in a way that can be understood by men and is focused on their specific needs, eg “what treatments are suitable for a 50year-old with localized cancer?” A website servicing this need would filter the information appropriately so that the visitor can quickly and easily access customized information. How a website functions in this way is an important aspect of its overall usefulness. Perhaps new studies need to ask, “How are men with prostate cancer actually finding information on the Internet?” and “Do the sites accessed deliver accurate, understandable information which meets their specific needs?” Carole Pinnock Urology Unit Repatriation General Hospital Daw Park, South Australia
Purpose: We assessed the quality of information available to patients on prostate cancer on the Internet. Materials and Methods: The search engine Webcrawler® was used with the search term “prostate cancer” to generate a list of 75 websites which were reviewed for currency, disclosure, attribution, interactivity and content. A rating tool was designed including 50 elements considered essential for a comprehensive review of prostate cancer, and each website was judged for degree of coverage and accuracy (each rated on a scale of 1 to 3) of information for each element. Results: Of the 75 sites 39 contained information about prostate cancer. Only 9 sites indicated a date of last update within 6 months. References were rarely given (in 5) and a disclaimer was provided on less than half of the sites (18). The sites covered a mean of 24 elements (range 6 to 43) with a mean coverage rating of 1.0 to 2.6 (1.8 overall). Of 943 elements covered on 39 sites, 94% were completely correct, 5% were mostly correct and 1% was mostly incorrect. Conclusions: The information on the Internet is of sufficient quality to aid in patient decision making. However, there are numerous shortcomings especially related to currency, disclosure and attribution. Degree of coverage is highly variable and there is a deficiency in balance of evidence found on many sites. The urologist needs to be aware of such shortcomings when counseling patients on prostate cancer. Key Words: prostatic neoplasms, internet
he quantity of information on the Internet related to prostate cancer is increasing rapidly, but the quality of this information has not been evaluated systematically.1 Concurrently, patients are becoming more facile with the Internet and are turning to it often as an important source of medical information. Up to 45% of patients with prostate cancer are using the Internet for self-education.1– 4 In some instances the information found on the Internet appears to be a key determinant in medical decision making.3,4 Information gathered from the Internet is usually not peer reviewed and can be posted without proper attribution.1 Many patients are probably not aware of these concerns.1,5 This is particularly problematic in prostate cancer, as it is a common malignancy associated with many controversial issues. The newly diagnosed patient is faced with numerous treatment options backed by minimal clinical evidence, further intensifying the need for useful information.6,7 In this study we assessed the quality of information available on prostate cancer on the Internet through a structured review of numerous websites. Specifically we wanted to determine if the quality of this information was adequate to support patient decision making.
T
Submitted for publication June 11, 2005. * Correspondence: Department of Urology, University of Washington, Box 356510, 1959 NE Pacific St., Seattle, Washington 98195 (telephone: 206-543-3640; FAX: 206-543-3272; e-mail: pblack@u. washington.edu). † Financial interest and/or other relationship with Sanofi-Aventis, Icos/Lilly, Amgen, Dendreon, Pfizer and Boehringer-Ingelheim.
See Editorial on page 1604.
0022-5347/06/1755-1836/0 THE JOURNAL OF UROLOGY® Copyright © 2006 by AMERICAN UROLOGICAL ASSOCIATION
MATERIALS AND METHODS An Internet search using the term “prostate cancer” without modifications or advanced search techniques was performed with the search engine Webcrawler®.3 Webcrawler® was chosen based on its search technology, which includes an evaluation of search results from other search engines. It reviews the information content of Web pages, their metatags, anchor text and link popularity to generate a search result that emphasizes relevance. The search yielded 390,770 websites, of which the top 75 sites were included in analysis. These websites were evaluated using a structured rating tool developed according to the “Criteria for Assessing the Quality of Health Information on the Internet” (1997)8 and standards and methodologies proposed by Bertrand et al for the systematic review of the quality of medical websites.9 Each website was evaluated by the research team for 5 central components, namely currency, disclosure, attribution, interactivity and content. The parameters describing the first 4 components are outlined in Appendix 1. Balance of evidence (Appendix 2) was evaluated as a component of attribution based on 4 controversial issues specifically related to prostate cancer: prostate cancer screening, watchful waiting vs immediate primary treatment, radiation treatment vs surgery and immediate vs delayed hormonal treatment. Between 3 and 5 elements of information were listed for each issue, and the websites were reviewed to evaluate the degree to which each statement was covered. The coverage scale outlined in table 1 was used. Content was measured with a rating tool designed to include 50 elements of information considered essential for a comprehensive review of prostate cancer. These elements
1836
Vol. 175, 1836-1842, May 2006 Printed in U.S.A. DOI:10.1016/S0022-5347(05)00996-1
PROSTATE CANCER INFORMATION ON INTERNET TABLE 1. Rating scales used to evaluate coverage and accuracy Description Coverage: 1 2 3 Accuracy: 1 2 3
Minimally addressed More than minimally addressed Mostly addressed Mostly incorrect Mostly correct Completely correct
1837
TABLE 2. Sites for which individual aspects of disclosure are present or absent
Name of organization Nature of organization Authorship Purpose of website Source of support
No. Indicated
No. Not Indicated
38 36 22 38 33
1 3 17 1 6
RESULTS
choosing radiation or surgery as a treatment modality for prostate cancer, but only 12 discussed at least 3 of the 5 rating elements more than minimally or completely. Five sites discussed the issue of immediate vs delayed hormone ablation, but only 2 in a balanced fashion. References were listed in a complete fashion on 5 sites, incompletely on 6 and not at all on 28. A disclaimer was provided on 18 of 39 sites. A contact or feedback mechanism was made available on 38 sites. The sites covered a mean of 24 elements (range 6 to 43) with 6 sites covering more than 35 elements (fig. 1). The mean coverage rating of individual sites ranged from 1.0 to 2.6 with an overall coverage rating of 1.8 (fig. 2). In total 943 elements were covered on the 39 sites, and of these 25% were “mostly addressed,” 34% were “more than minimally addressed” and 40% were “minimally addressed.” The mean total coverage score was 45 out of a possible 150. The scores ranged from 6 to 88 with 9 sites scoring 1 to 25, 14 sites scoring 26 to 50, 13 sites scoring 51 to 75 and 3 sites scoring greater than 75. Figure 3 shows how the scores correlated to the type of site. Each element was covered on a mean of 19 sites (range 4 to 34). Categorizing the elements revealed that the elements of the given categories were covered on a mean of 18.8 to 24.1 sites for all categories except hormone refractory disease, for which the elements were covered at least minimally on a mean of 8.7 sites (fig. 4). Comparing coverage scores for elements categorized by modality of treatment (radical prostatectomy, external beam radiation, brachytherapy and endocrine treatment), a statement can be made about the balance with which these categories are covered on any given site. A total of 33 sites
The 75 sites included 59 original sites and 16 dead links. Only 39 of the 59 sites had information about prostate cancer, and these 39 sites were evaluated in detail. Six sites were managed by academic institutions or government organizations, 11 were nonprofit organizations, 12 were commercial, 4 were sponsored by individual physicians or private health care organizations, 4 were personal sites including self-help groups, and 2 were news sites. A date of last update was not indicated in 24 of 39 sites. One site had been updated within 2 weeks, 4 within 1 month and 4 within 3 months. Six sites had been updated more than 6 months before the date of review. The data on disclosure are summarized in table 2. Balance of evidence was evaluated using the rating tool designed for this purpose. A total of 25 sites discussed prostate cancer screening, but only 10 described at least 2 of the 4 elements more than minimally or completely. There were 24 sites that discussed watchful waiting and 22 discussed at least 2 of the 3 rating elements more than minimally or completely. The 25 sites addressed the decision process for
FIG. 1. Number of sites that cover specified range of elements. Mean number of elements per site is 24, with individual elements covered on range of 6 to 43 sites.
were derived from the corresponding chapters in Campbell’s Urology10 and are listed in Appendix 3. A minimum of 5 elements had to be covered for a site to be included in the survey. The categories of elements included epidemiology/ etiology/natural history, diagnosis/staging/pathology, radical prostatectomy, external beam radiation, brachytherapy, endocrine treatment, hormone refractory prostate cancer and alternate treatments. Each website was judged for degree of coverage and accuracy of information for each element found on the website using the scales described in table 1. The raw data for coverage generated with this rating tool were analyzed with 3 methods. The number of elements covered by each website was tabulated. Then the mean rating on the scale of 1 to 3 for all elements covered on a site was calculated. Finally, a score was calculated for each website that consisted of the sum of the ratings for each individual element covered. The maximum possible score for 50 elements rated on a scale of 1 to 3 is 150. The scores obtained for each category of elements within a single website could be compared by adding together the ratings within a category to get a sum for that category. The sums could not be compared directly between categories because each category had a different number of elements. Therefore, each score was expressed as a percentage (percentage score) of the total possible score for that category, and the percentages were compared between categories, providing further information on the balance of evidence given on that website.
1838
PROSTATE CANCER INFORMATION ON INTERNET
FIG. 2. Number of sites with mean degrees of coverage falling in specified limits. Overall mean was 1.8, with individual sites ranging from 1.0 to 2.6.
described at least 1 of the 4 standard treatment modalities with at least 20% of the maximum coverage score for that category. Of these 33 sites 1 site described only 1 of the 4 treatment modalities (brachytherapy), while 7 sites left out at least 1 treatment modality completely (1 radical prostatectomy, 2 brachytherapy, 5 endocrine treatment and 1 site left out 2 modalities). If only sites that discussed all 4 treatment modalities are included (33), there were discrepancies in the amount of information provided on each modality. An imbalance of evidence, as indicated by a 20% difference (more or less) in the percentage scores between 1 category and the next closest category, was present in 13 of 33 sites. This consisted of overrepresentation or underrepresentation of radical prostatectomy on 2, external beam radiation on 2, brachytherapy on 3 and endocrine treatment on 5 sites. Of 943 elements covered in 39 sites, 94% were “completely correct,” 5% were “mostly correct” and 1% was “mostly incorrect.” The mean accuracy rating achieved was 3 in 17 sites, 2.9 in 15 sites, 2.8 in 5 sites and 2.7 or 2.6 in 1 site each. The overall accuracy rating was 2.9. The most common
FIG. 3. Median coverage score for each type of website. Mean score for all sites was 45, range 6 to 88. univ/gov, university and government. comm, commercial.
FIG. 4. Mean number of sites covering elements of each category. Number ranges from 8.7 for hormone refractory disease (HR) to 24.1 for diagnosis and staging (diag/stag). epid/etiol, epidemiology and etiology. EBXRT, external beam radiation. brachy, brachytherapy. endo, endocrine treatment.
points of error were the definition of Gleason sum (defined completely in 27 sites, with 20 “completely correct” and 7 “mostly correct”) and the guidelines for PSA screening (defined completely in 16 sites, with 12 “completely correct” and 4 “mostly correct”). Otherwise all ratings for “mostly incorrect” and “mostly correct” were scattered evenly among the 50 elements. Accuracy did not differ by type of site. DISCUSSION This evaluation of the quality of information on prostate cancer on the Internet reveals a variety of important findings related to currency, disclosure, attribution, interactivity and content.11–13 Approximately two-thirds of sites did not indicate the currency of the information presented on the site. This would appear to be a fundamental oversight in any biomedical medium, as the information content changes rapidly. Only 9 sites had been updated within 6 months, which would be an appropriate interval for this kind of information.11 Disclosure is important on web pages because it allows the reader to judge the intent and objectivity of the information, as well as its source. Although the name and nature of the organization constructing the website, and the purpose of the website and the source of sponsorship were clearly indicated in most sites, actual authorship was listed on just more than half of sites. Attribution allows the reader to assess the quality of the information being provided. While a list of references is never absent from a peer reviewed publication, only 5 sites included a complete list of references, and 28 had no references. Although most patients may not pursue further references, it is essential to give the reader the opportunity to verify independently the information offered.14 Similarly, a description of the validity of information (level of evidence) was addressed on only a small number of sites (7 of 39), and in none with any sort of completeness. Every site that could potentially be used for patient education should be expected to include a disclaimer that a patient needs to discuss any newly found information with a health care provider to
PROSTATE CANCER INFORMATION ON INTERNET apply it to his or her individual case, yet only half of the sites in the current study did this. Balance of evidence is a more difficult parameter to assess, but it may be the most important qualitative parameter after content. The results of the current study indicate that there is an adequate balance across all sites (as measured by the number of sites covering elements within the different categories), except information on hormone refractory disease, but that there is a significant imbalance of evidence within individual websites. The coverage of the 4 primary treatment modalities (radical prostatectomy, external beam radiation, brachytherapy and endocrine treatment) was poorly balanced on approximately a third of sites. Endocrine treatment was overrepresented or underrepresented more often than the other modalities, but for the most part the oversights were equally distributed among the 4 modalities. The imbalance was further evident in the coverage of 4 controversial yet central questions on prostate cancer diagnosis and management. One issue was presented in a balanced fashion on only 2 sites, 2 others on a third of sites and the fourth on two-thirds of sites. The issue of immediate vs delayed hormone ablation proved to be a poor example for measuring balance of evidence, since only a small number of sites addressed this topic. However, screening for prostate cancer, the advantages and disadvantages of watchful waiting, and a comparative evaluation of surgery vs radiation treatment are key issues that are not covered in a balanced fashion. The ultimate factor in assessing a website’s quality is its content, which can be looked at as coverage and accuracy. If the content is deficient, the other parameters will not make it a useful site. Hellawell et al have also looked specifically at the information provided on prostate cancer on the Internet and attempted to quantify it. They scored websites on a scale from 10 to 100 and found a mean score of 50.7.1 In the current study while some sites provide excellent coverage of many facets of the disease, others are limited in scope. Six sites were clearly superior to all others with respect to the amount of coverage (more than 35 of the 50 elements). Although the risk of such exhaustive coverage is to overwhelm the reader, in general these sites are carefully constructed and divided into sections that allow easy browsing through relevant information and omission of topics less pertinent to the reader. Interestingly, the top 6 sites varied from public service sites such as the Australian Lions Society and selfhelp groups like UsTOO, to government institutions such as the National Cancer Institute and industry sponsored sites such as Prostatecancer.com (AstraZeneca). However, the percentage scores, another measure of coverage, were higher for the sites sponsored by universities/government and nonprofit organizations compared to those sponsored by private physicians/hospitals and commercial sites. The 8 sites that covered less than 15 elements were mostly sites promoting a specific aspect of prostate cancer or promoting a product. Accuracy of information was excellent across most websites with 99% of information being “mostly” or “completely correct.” Only 6 sites had an accuracy rating of less than 2.9 (best possible 3.0). Although these data on accuracy are for the most part reassuring, the goal should be to avoid any report of false information. In addition, high accuracy is tempered by poor balance, which may be considered equivalent to error of omission15 and can generate a misleading impression.
1839
The main limitation of the current study is the lack of validity of the survey tool and the subjectivity of content elements. We have chosen one search engine from many available, and we have evaluated a relatively small number of sites out of the overall number of sites available. Since these include the most frequently visited sites, we believe that they are representative of what patients are finding. In addition, there are aspects of website design such as readability and user-friendliness that we have not attempted to define, but these also will have a significant impact on the usefulness of a website to an individual patient. We have made no attempt to define which patients are using the Internet and to what purpose and, therefore, are not able to predict what information should be offered and how it should be best presented. Other authors have addressed the question of what could be done to improve the quality of information on the Internet. Pautler et al suggest that national organizations such as the American Urological Association develop a list of approved/endorsed websites, that guidelines be developed for the construction of balanced and responsible websites, and that physicians learn to guide their patients to appropriate websites.3 The American Medical Association offers guidelines for website design16 and a system of website recognition has been developed by organizations such as HONcode.17 Although the quality of Internet information is important, physician familiarity with these data is equally important. Physicians continue to have the greatest influence on patient decision making,7 and it would become us to use the Internet as a counseling tool for suitable patients. CONCLUSIONS The information on the Internet is of sufficient quality to aid in patient decision making. However, there are numerous shortcomings especially related to currency, disclosure and attribution. Degree of coverage is highly variable but the covered content is accurate. The most significant shortcoming is the deficient balance of evidence found on many sites. The urologist needs to be aware of such shortcomings when counseling patients on prostate cancer, and should be able to guide the patients to up-to-date websites that offer balanced information from identifiable, reputable sources. APPENDIX 1 Summary of parameters used to rate individual websites Parameter Currency/date of last update Disclosure: Name of organization Nature of organization Authorship Purpose of website Source of support Attribution: Balance of evidence List of references Validity of information Disclaimer Interactivity: Contact/feedback mechanism Chat room/bulletin boards
Rating Less than 2 wks/less than 1 mo/less than 3 mos/less than 6 mos/greater than 6 mos/not indicated Present/absent Present/absent Present/absent Present/absent Present/absent See Appendix 1 None/incomplete/complete None/incomplete/complete Present/absent None/e-mail/telephone/FAX/ mailing address Present/absent
1840
PROSTATE CANCER INFORMATION ON INTERNET
APPENDIX 2 Rating Tool for Balance of Evidence BALANCE OF EVIDENCE 1. Screening for prostate cancer ● annual DRE and PSA recommended for males over 50 years old; for black males and other high risk males over 45 years old ● no randomized studies have been completed to show an advantage in overall survival by screening with PSA and DRE ● morbidity and cost of tests, including PSA and prostate needle biopsies ● there is no evidence that early aggressive treatment reduces mortality 2. “Watch and wait” versus immediate treatment ● if low grade (Gleason score), low volume disease, low clinical stage ● if patient older, more comorbidities ● then may not need treatment; more likely to die from other cause, may suffer adverse effects of treatment without experiencing benefits of treatment 3. Radiation vs brachytherapy vs radical prostatectomy for organ confined disease ● difficult to compare outcomes of radiation and surgery, because radiation cases tend to be a selection of cases that are not surgical candidates, and there is no pathological staging (pT2 vs pT3) available in radiation cases ● there are no good randomized controlled prospective data comparing different forms of treatment; retrospective studies indicate similar local control between brachytherapy and radical prostatectomy in low stage disease after 5 and 10 years; poorer local control after EBRT ● longer followup pending for radiation/brachytherapy ● each has advantages and disadvantages regarding tolerability and adverse effects (list 2 major adverse effects for each treatment) ● “best” treatment needs to be decided in each patient individually 4. Immediate vs delayed hormone treatment ● only 1 study clearly shows benefit to early hormone treatment; this is in patients with positive lymph nodes found during radical prostatectomy ● otherwise no study shows definite benefit with early vs late hormones (ie decreased complication rate or improved survival) ● hormone treatment involves significant adverse effects and possible decrease in quality of life
APPENDIX 3 Rating Tool for Content
EPIDEMIOLOGY/ETIOLOGY 1999: approximately 179,500 new cases, 37,000 deaths; second leading cause of cancer death in men (higher number earlier in the 90s eg greater than 300,000 cases, greater than 40,000 deaths in 1996) demographics: 95% between age 45 and 89, median age 72 highest incidence in blacks, intermediate in whites, lowest in Chinese/Japanese family history: higher risk if higher number of affected relatives and if affected relatives have earlier age of onset some (eg American Urological Association and American Cancer Society) recommend annual DRE and PSA starting at age 50, others (eg Canadian Task Force on the Periodic Health Examination and the United States Preventive Service Task Force) do not initial studies indicate that selenium and vitamin E reduce the incidence of prostate cancer compared to a placebo; a large national trial is currently underway to verify this PATHOLOGY high grade PIN on needle biopsy approximately 30% to 50% risk of finding carcinoma on subsequent biopsies; similar with atypical small acinar proliferation; repeat biopsy should be performed for both entities definition of Gleason grading: based on glandular pattern of the tumor as identified on relatively low magnification; the primary and secondary architectural patterns are identified and assigned a grade 1 to 5 (5 is least differentiated); sum of both is Gleason score, this is an indicator of how aggressive the tumor is DIAGNOSIS/STAGING usually asymptomatic; sometimes associated with voiding symptoms in locally advanced disease and bony pain in metastatic disease, also acute spinal cord compression; some cancers are diagnosed incidentally at the time of TURP screening: most effective method is PSA plus DRE; 18% will be found by DRE, 45% by PSA and 37% by PSA and DRE together diagnostic tests: TRUS: classically shows hypoechoic lesion in periphery, but carcinoma may be hypo-, iso- or hyperechoic; main use of TRUS is for biopsy guidance needle biopsy: classically sextant biopsy: base, mid and apex bilaterally; newer evidence indicates that 8 to 12 may be better Staging Systems Whitmore-Jewett staging system A1 focal/A2 diffuse CaP on TURP/simple prostatectomy B1 palpable but organ confined unilaterally/B2 bilaterally C1 minimal capsular invasion/C2 extensive capsular invasion with hydro or bladder outlet obstruction D1 lymph node mets/D2 distant mets/D3 hormone refractory Clinical TNM staging system T1 nonpalpable ⫽ normal DRE T1a 5% or less/T1b greater than 5% of specimen on TURP/simple prostatectomy T1c by PSA only T2 palpable, organ confined T2a unilateral less than half/T2b unilateral greater than half/T2c bilateral T3 beyond capsule T3a extracapsular CaP/T3b seminal vesicle invasion T4 invasion of bladder neck, external sphincter, levator, rectum, pelvic wall N0 no nodes/N1 regional nodes M1a nonregional nodal mets/M1b bone/M1c visceral mets Staging: pelvic imaging with computerized tomography or magnetic resonance imaging for detection of local extension or presence of lymph node metastases is not routinely useful because of low sensitivity; possible exception in men with locally advanced disease by DRE, PSA greater than 20 or Gleason 8 to 10 bone scan indicated for patients with PSA greater than 20; bone metastases are typically osteoblastic (appendix 3 continued)
PROSTATE CANCER INFORMATION ON INTERNET
1841
APPENDIX 3 continued
NATURAL HISTORY prospective randomized controlled trials investigating the effect of early detection and treatment of prostate cancer on survival have not yet been completed Survival: approximately 20% of men with CaP will die of disease median survival of men who have: - PSA recurrence after local treatment approximately greater than 10 years - metastatic disease approximately 2–3 years - hormone refractory disease approximately 12–18 months watch and wait: some prostate cancers are clinically insignificant and do not require treatment; these can be watched with regular H&P and PSA surveillance; the patient is followed until either symptoms or an increase in PSA leads to treatment RADICAL PROSTATECTOMY (RRP) usual indication: life expectancy greater than 10 years, clinical T1–2 disease; high PSA (eg greater than 20) is not a contraindication by itself as long as no other evidence of metastatic disease PSA is used to follow after RRP - should be undetectable; probability of PSA recurrence correlates to preoperative PSA, Gleason score (in needle biopsy and in final surgical specimen) and pathological stage 5 year progression-free survival rate for T1–2, Nx disease is 69% to 83%; 10 year progression-free survival rate 47% to 78%; greater than 90% progression-free survival for organ confined disease invasion of seminal vesicles and involvement of pelvic lymph nodes are poor prognostic indicators; significance of positive margins is uncertain - only 40% to 50% of patients with positive margins (and no other adjuvant treatment) will have PSA recurrence within 5 years complications: most common early complication is hemorrhage; average blood loss 800 to 1,500 ml operative mortality approximately 0.5% (death within 30 days of RRP) also: rectal injury in less than 1%, wound infection/dehiscence in 1.5%, lymphocele in 3% to 4%, anastomotic stricture in 0.5% to 9.0%, DVT in 1% to 2%, PE in 1% to 2% rate of postoperative incontinence depends on source of data: patient surveys give rate of 19% to 31%, centers of excellence report 5% to 10%; details of surgical technique are thought to be important in determining rate of continence; can improve up to 24 months postoperatively the cavernosal nerves, which are essential for erections, are resected during a standard RRP; in a nerve sparing prostatectomy they can be saved on 1 or both sides of the prostate if this is thought not to jeopardize tumor control return of erectile function depends on potency status prior to RRP, stage of disease, unilateral vs bilateral nerve sparing and age of patient; up to 75% to 80% of previously potent men younger than 60 years with bilateral nerve sparing will retain erectile function; population surveys give lower numbers than are reported from centers of excellence; can improve up to 24 months postoperatively RADIATION-EXTERNAL BEAM alternative to prostatectomy and brachytherapy in T1–2 disease; it has not been definitively shown that survival after external beam radiation is better or worse than prostatectomy or brachytherapy treatment of choice, combined with hormone ablation (for cytoreduction), for T3 disease, but poor progression-free survival rates at 5 and 10 years (15% to 30%) external beam radiation can be offered adjuvantly to radical prostatectomy in three settings: 1. occult residual tumor in pT3 disease, 2. PSA recurrence after RRP with extraprostatic extension, and 3. clinical local failure; should wait 5 to 8 weeks to allow for return of continence and complete healing from RRP most centers use “conventional simulation”; however, newer “3-dimensional conformal delivery” reduces complications (computer mapping of tumor within prostate, patient fitted with plastic mold resembling a cast to limit motion, improve accuracy of aim) short-term adverse effects include diarrhea, rectal irritation, dysuria, frequency chronic complications in 5%: rectal/bladder injury (including rectal bleeding and hematuria), enteritis, urethral stricture, 0.9% incontinence, 40% to 70% impotence RADIATION-BRACHYTHERAPY TRUS guided placement of radioactive I125 or Pd103 into prostate through the perineum; seeds emit radioactivity for 6 (Pd103) to 12 (I125) months; conceptual advantage: higher radiation dose to prostate with less exposure to surrounding tissue; dose greater than 100 Gy can be offered to any patient with clinically localized prostate cancer as an alternative to prostatectomy or external beam radiation; relative contraindications include large prostate, severe obstructive voiding symptoms and s/p TURP combined with EBRT in patients with high risk for extracapsular disease (T2b–T3a, Gleason 7 to 10, PSA greater than 10) short-term results (PSA-free survival) similar to radical prostatectomy, and better than EBRT for organ confined disease; long-term results pending short-term adverse effects include acute urethral irritiation, urinary obstruction (including retention) and proctitis long-term adverse effects include urinary incontinence (higher after TURP), diarrhea and erectile dysfunction (especially older patients); these problems may increase with time popular perception is that brachytherapy causes less impairment of quality of life, but randomized studies are pending; appears to cause less incontinence and impotence but more irritative symptoms ENDOCRINE TREATMENT includes orchiectomy, LH-RH agonists (mainstay), antiandrogens and estrogens; no regimen is superior to any other in survival indication: locally advanced or metastatic disease; also for patients with high grade, organ confined disease who are not candidates for other treatments and patients with PSA recurrence after local treatment androgens are made in the testes and adrenal glands (5% to 10%); only testicular production ceases with LH-RH agonist or orchiectomy; total androgen blockade is LH-RH agonsist or surgical castration plus antiandrogen concurrently; the antiandrogen blocks the effect of testosterone made extratesticularly; there is no evidence that adding an antiandrogen improves time to progression or survival side effects include loss of libido and potency (but not in all men); also hot flashes, osteoporosis, fatigue, loss of muscle mass, anemia, weight gain, elevated liver enzymes; libido and potency maintained in pure antiandrogen therapy intermittent endocrine treatment: theoretical advantage of prolonging survival by delaying progression to androgen independence – subject of ongoing randomized studies initial studies indicate that there is an improvement in quality of life, including sexual function, during the off-therapy interval a period of 6 to 9 months on therapy is usually recommended; the mean off-therapy interval approaches 50% of the duration of the treatment cycle neoadjuvant endocrine treatment: there appears to be no advantage (biochemical recurrence) to 3 months of neoadjuvant hormone ablation; studies are ongoing looking at 8 months of treatment; the rate of positive surgical margins is lower after neoadjuvant endocrine treatment, but the clinical significance is uncertain TREATMENT OF HORMONE REFRACTORY CARCINOMA testosterone level should be checked to evaluate adequacy of hormone ablation - should be castrate levels; hormone ablation is continued during chemotherapy antiandrogen withdrawal can result in temporary clinical response seen as decreased PSA, symptomatic relief; requires serial monitoring of PSA every 4 to 8 weeks; median duration of effect approximately 3 months (appendix 3 continued)
1842
PROSTATE CANCER INFORMATION ON INTERNET
APPENDIX 3 continued
second-line endocrine intervention includes ketoconazole, diethylstilbestrol, corticosteroids or aminoglutethimide cytotoxic agents used include mitoxantrone, prednisone, estramustin, taxol, taxotere, cyclophosphamide, etoposide, vinblastine, doxorubicin first line palliative treatment for symptomatic bone lesions is external beam radiation; strontium 89 can increase symptom-free interval by targeting subclinical disease if neurological deficits, first line treatment is high dose corticosteroids, then external beam radiation; surgery if unstable fracture, progression during radiation or recurrence after radiation future directions of chemotherapy include topoisomerase I inhibitors (TOPO-I), tissue inhibitors of metalloproteases (TIMPS), inhibition of angiogenesis, differentiation agents (eg, Vit D, butyrates), gene therapy and monoclonal antibodies ALTERNATIVE TREATMENTS cryoablation is a rarely performed treatment, accepted currently only for use after local radiation failure; uses probes placed by TRUS guidance; urethral warming catheter to protect from necrosis; advancement of ice ball observed by TRUS with 1 to 2 mm accuracy; extends up to muscularis propria of rectum salvage prostatectomy: used for local recurrence following radiation in select patients; defined as positive biopsy in conjunction with rising PSA and no evidence of metastases; preoperative PSA should be less than 10, (15% vs 86% advanced pathological features)
Each element (first level bullet) was judged for coverage and accuracy according to the scales outlined in table 2.
Abbreviations and Acronyms CaP DRE EBRT H&P LH-RH PSA RRP TRUS TURP
⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽ ⫽
prostate cancer digital rectal examination external beam radiation therapy history and physical examination luteinizing hormone releasing hormone prostate specific antigen radical retropubic prostatectomy transrectal ultrasound transurethral resection of the prostate
REFERENCES 1. Hellawell, G. O., Turner, K. J., Le Monnier, K. J. and Brewster, S. F.: Urology and the Internet: an evaluation of internet use by urology patients and of information available on urological topics. BJU Int, 86: 191, 2000 2. Diefenbach, M. A., Dorsey, J., Uzzo, R. G., Hanks, G. E., Greenberg, R. E., Horwitz, E. et al: Decision-making strategies for patients with localized prostate cancer. Semin Urol Oncol, 20: 55, 2002 3. Pautler, S. E., Tan, J. K., Dugas, G. R., Pus, N., Ferri, M., Hardie, W. R. et al: Use of the internet for self-education by patients with prostate cancer. Urology, 57: 230, 2001 4. Smith, R. P., Devine, P., Jones, H., DeNittis, A., Whittington, R. and Metz, J. M.: Internet use by patients with prostate cancer undergoing radiotherapy. Urology, 62: 273, 2003 5. Henson, D. E.: Cancer and the Internet. Cancer, 86: 373, 1999 6. Berry, D. L., Ellis, W. J., Woods, N. F., Schwien, C., Mullen, K. H. and Yang, C.: Treatment decision-making by men with localized prostate cancer: the influence of personal factors. Urol Oncol, 21: 93, 2003 7. Patel, H. R. H., Mirsadraee, S. and Emberton, M.: The patient’s dilemma: prostate cancer treatment choices. J Urol, 169: 828, 2003 8. http://www.mitretek.org/hiti/showcase/documents/criteria.html. Accessed April 4, 2002 9. Berland, G. K., Elliott, M. N., Morales, L. S., Algazy, J. I., Kravitz, R. L., Broder, M. S. et al: Health information on the Internet: accessibility, quality, and readability in English and Spanish. JAMA, 285: 2612, 2001 10. Walsh, P. C., Retik, A. B., Vaughan, E. D., Jr. and Wein, A. J.: Campbell’s Urology, 7th ed. Philadelphia: W. B. Saunders Co., 1998
11. Al-Bahrani, A. and Plusa, S.: The quality of patient-orientated internet information on colorectal cancer. Colorectal Dis, 6: 323, 2004 12. Gilliam, A. D., Speake, W. J., Scholefield, J. H. and Beckingham, I. J.: Finding the best from the rest: evaluation of the quality of patient information on the Internet. Ann R Coll Surg Engl, 85: 44, 2003 13. Lee, C. T., Smith, C. A., Hall, J. M., Waters, W. B. and Biermann, J. S.: Bladder cancer facts: accuracy of information on the Internet. J Urol, 170: 1756, 2003 14. Darmoni, S. J., Amsallem, E., Haugh, M., Lukacs, B., Leroux, V., Thirion, B. et al: Level of evidence as a future gold standard for the content quality of health resources on the internet. Methods Inf Med, 42: 220, 2003 15. Fagerlin, A., Rovner, D., Stableford, S., Jentoft, C., Wei, J. T. and Holmes-Rovner, M.: Patient education materials about the treatment of early-stage prostate cancer: a critical review. Ann Intern Med, 140: 721, 2004 16. http://pubs.ama-assn.org/how_principle.htm. Accessed July 8, 2004 17. http://www.hon.org. Accessed March 25, 2003
EDITORIAL COMMENT The authors present a valuable study on content accuracy of prostate cancer websites using Campbell’s Urology as a gold standard. However, there is a question of whether the average male searching for prostate cancer information would retrieve the same 39 sites reviewed here out of a possible 390,000. Moreover, information needs to be delivered in a way that can be understood by men and is focused on their specific needs, eg “what treatments are suitable for a 50year-old with localized cancer?” A website servicing this need would filter the information appropriately so that the visitor can quickly and easily access customized information. How a website functions in this way is an important aspect of its overall usefulness. Perhaps new studies need to ask, “How are men with prostate cancer actually finding information on the Internet?” and “Do the sites accessed deliver accurate, understandable information which meets their specific needs?” Carole Pinnock Urology Unit Repatriation General Hospital Daw Park, South Australia