Female fertility in the cancer setting: availability and quality of online health information

Abstract

As cancer survival rates continue to increase, so does the relevance of starting or expanding a family after cancer, impacting quality of life. However, the option of fertility preservation is not always addressed by physicians before treatment. In the meanwhile, many patients look for health information online. Although potentially helpful in the decision-making process, the Internet could also have adverse effects for this vulnerable patient group when information is incorrect. We aimed to assess the availability and quality of information on female oncofertility on websites of (inter)national oncology, fertility and oncofertility organizations. Availability was identified using an original checklist. Quality was measured by reliability (DISCERN and LIDA instruments), usability (LIDA instrument), and readability (Flesch Kincaid Grade Level test). Of all investigated websites, 52% contained information on oncofertility, with 32% containing a stand-alone page. Of those sites, the amount of treatment options covered were equivalent between specialty sites and general fertility and oncology sites (p = 0.07). The quality of information on treatment choices was the highest for oncofertility websites. Reliability and usability did not differ significantly between websites. No websites evaluated were at the recommended 7th–8th grade reading level. Availability and quality should be improved, and high-quality resources are recommended by physicians.

Keywords:

• Fertility preservation
• cancer
• communication

Introduction

In female cancer patients, post-treatment quality of life becomes increasingly important, as survival rates increase (Deshpande, Braun, & Meyer, 2015). Fertility plays a major role in this quality of life, particularly for patients of reproductive age (Deshpande et al., 2015; Goncalves & Quinn, 2016). The American Society for Clinical Oncology (ASCO) encourages physicians to advise patients on potential threats to their fertility as early as possible in the treatment process, so as to allow for the widest range of options for fertility preservation, although the focus in treatment may be initially on the cancer diagnosis itself (Loren et al., 2013).

In this digital age, patients not only learn from traditional physician instructions but also frequently consult the Internet as a resource for information on their diagnosis, treatment options and prognosis (Higgins, Sixsmith, Barry, & Domegan, 2011; Tan & Goonawardene, 2017). Interest in, and utilization of, the Internet as a resource for health information is growing tremendously (Higgins et al., 2011). About 70% of internet users in the United States and Europe seek health information online, and the most active users are those aged 30 to 44 (Higgins et al., 2011; ‘Pew Research Center. Health Fact Sheet,’ 2013), in particular, women (Higgins et al., 2011). Additionally, 20% of doctors refer their patients to websites (Podichetty, Booher, Whitfield, & Biscup, 2006). Cancer patients might even be more likely to turn to the Internet for information, since the option of fertility preservation is not always discussed. For example, 25% of breast surgeons discussed fertility only if mentioned by their young female breast cancer patients (Warner et al., 2016). As a result, those patients who are not referred are left to locate this vital information on their own. This is critically important because they may be in great distress, and, therefore, be particularly vulnerable to locating incorrect information. Moreover, in 29% of women with breast cancer, decisions are influenced by infertility concerns (Partridge et al., 2004). A significant advantage of the internet is the ability to aid in difficult decision-making processes at a time when choices must be made quickly if the information on the Internet is authoritative and reliable (Ehrbar et al., 2018; Garvelink et al., 2013; Hershberger, Finnegan, Altfeld, Lake, & Hirshfeld-Cytron, 2013; Muller et al., 2017).

It has been concluded that the quality of online information is highly variable (Berland et al., 2001; Eysenbach, Powell, Kuss, & Sa, 2002; Storino et al., 2016). Absent or incorrect information on websites could limit women in seeking appropriate care. Therefore, we aimed to evaluate the availability and quality of oncofertility information for female cancer patients online.

Materials and methods

Data collection

First of all, we wanted to identify websites of oncology, fertility and oncofertility organizations, since this is where we expect cancer patients to search and want to guarantee high quality information; 27% of internet users begin their investigations with a specific health-related website (Higgins et al., 2011). To identify these websites, a search was conducted in May 2016 (Figure 1). The website CancerIndex (‘CancerIndex.org’) was used to identify a list of oncology organizations, since this was the most comprehensive list found online. Oncofertility and fertility organizations were identified using the Oncofertility Consortium Global Partners Network, members of the International Federation of Fertility Societies (IFFS) (‘International Federation of Fertility Societies. Membership page’), and by mining the references and resources listed on these organizations’ websites. Additionally, a Google search was performed to identify oncofertility organizations, using the search terms ‘oncofertility’ OR ‘fertility preservation’ (June 2016). Websites were subdivided into affiliation (professional, patient advocacy, governmental) categories (definition according to the National Health Council (USA)). Websites had to meet the following inclusion criteria: (i) contain health information for adult female cancer patients (aged 18–40); (ii) be from (inter)national organizations based in countries that offer fertility preservation; and (iii) written in English. For oncology organizations, we only included the incidence-top-5 cancers per World Health Organization region using the GLOBOCAN 2012 online analysis (Ferlay et al., 2013) (see Supplementary on-line material). To identify the countries that offer fertility preservation, we used the 61 member nations of the International Society for Fertility Preservation, Asian Society for Fertility Preservation, Oncofertility Consortium, Latin America Oncofertility Consortium, NORDFERTIL, FertiPROTEKT and FuTuRE Fertility. We excluded local organizations, and websites solely aimed at health care professionals, children or men. The initial screening was conducted by AdM and a part of the Google search by VA. In the qualification process, two raters were assigned to independently qualify each website in a pool of four raters (AR, AdM – see author list; and VA, SA – see Acknowledgements).

Figure 1. PRISMA flow diagram search method.

Assessment of content

The websites were assessed for oncofertility content using an original checklist (see supplementary on-line material) with 15 items, including presence of a Health on the Net Code of Conduct (HONcode) certification. The HONcode is an ethical standard aimed at offering good quality health information (Boyer, Selby, Scherrer, & Appel, 1998). Only websites that provided a stand-alone page on oncofertility were further qualified with items 6–15.

Assessment of quality

Websites were then qualified using validated instruments, including the DISCERN instrument, LIDA instrument and a readability test. The DISCERN instrument measures reliability, evaluates the quality of information on treatment choices and provides an overall rating (Charnock, Shepperd, Needham, & Gann, 1999). The LIDA instrument (Version 1.2, Minervation Ltd, Oxford, UK) measures the accessibility and reliability of the website (‘The Minervation validation instrument for healthcare websites (LIDA instrument, 2007). For both instrument rates, percentages were calculated with scores <50% considered low, 50–90% medium and any >90% high (Doğramacı & Rossi-Fedele, 2016; Rossi-Fedele, Musu, Cotti, & Doğramacı, 2016). The Flesch Kincaid Grade Level (FKGL) test is a readability test measuring the average number of syllables per word and the average number of words per sentence, rating on a U.S. school grade level (Kincaid, Fishburne, Rogers, & Chissom, 1975). A 7th–8th grade reading level is recommended for web content (‘National Institutes of Health. How to Write Easy to Read Health Materials (2016). National Library of Medicine Web site’).

Statistical analyses

Data were treated as nonparametric. The items from the checklist answered with ‘yes’ or ‘no’ responses were treated as dichotomous variables, a Fisher's exact test was run due to small, expected frequencies (Fleiss, 2003). Checklist items 6 and 14 answered with a number, as well as the DISCERN, LIDA and FKGL, were treated as continuous variables and analyzed by use of the Kruskal–Wallis H test. Statistical significance was set at p ≤ 0.05. Pairwise comparisons were performed using Dunn’s (1964) procedure with a Bonferroni correction for multiple comparisons, where statistical significance was accepted at the p < 0.0083 level. Scores for DISCERN and LIDA instruments were analyzed for inter-rater reliability using the intra-class correlation (ICC) with a two-way mixed model for absolute agreement, indicating moderate scores for DISCERN (0.73 (95% CI, 0.52–0.86)) and LIDA (0.8 (95% CI, 0.64–0.90)) (McCool, Wahl, Schlecht, & Apfelbacher, 2015; Shrout, 1998). Statistical analyses were performed using IBM SPSS Statistics software 24 (SPSS, Chicago, IL).

Results

The search identified a total of 1811 oncology, fertility and oncofertility organizations. After applying eligibility criteria, a total of 104 websites were included in the final analysis: 69 from oncology, 28 from fertility and seven from oncofertility websites. When subdivided into affiliation, the sites broke down as 33 professional organizations, 69 advocacy organizations and two governmental organizations (see Supplementary on-line material eTable 1^{Table 1}). The majority of websites found were from organizations located in North America (38% (39/104)), followed by Asia (23% (24/104)), Europe (19% (20/104)), Oceania (13% (13/104)), international organizations (5% (5/104)) and Africa (3% (3/104)). No English language-based websites from Latin America and the Caribbean could be found.

Table 1. Website content: availability of basic information on oncofertility^a.

Items	All websites (n = 103)^b	Oncology (n = 69)	Fertility (n = 27)^c	Oncofertility (n = 7)	p Value	Professional (n = 32)^d	Patient (advocacy) (n = 69)	Governmental (n = 2)	p Value
1. Full-access	103 (99)	69 (100)	27 (96)	7 (100)	0.34	32 (97)	69 (100)	2 (100)	0.34
2. Effect of cancer (treatment) on fertility	70 (68)	46 (67)	17 (63)	7 (100)	0.16	23 (72)	45 (65)	2 (100)	0.62
3. Female specific information	66 (64)	43 (62)	16 (59)	7 (100)	0.12	22 (69)	42 (61)	2 (100)	0.51
4. Fertility preservation option	54 (52)	35 (51)	12 (44)	7 (100)	0.02	15 (47)	37 (54)	2 (100)	0.51
5. Oncofertility	33 (32)	19 (28)	7 (26)	7 (100)	<0.001	10 (31)	21 (30)	2 (100)	0.18

Key.

^a Data presented as number (%). Websites divided by oncology, fertility and oncofertility organizations as well as affiliation.

^b For I1, n = 104; for I2–5, n = 103.

^c For I1, n = 28; for I2–5, n = 27.

^d For I1, n = 33; for I2–5, n = 32.

Website content: availability of basic information on oncofertility

Of all 103 fully accessible websites, the effects of cancer treatment on fertility were mentioned by 68% (70/103). More specifically, the effects were mentioned by 67% (46/69) of oncology websites, 63% (17/27) of fertility organizations and 100% (7/7) of oncofertility organizations (p = 0.16) (Table 1). The option of fertility preservation was mentioned by 52% of sites (54/103), which is 51% (35/69) of oncology websites, 44% (12/27) of fertility websites and 100% (7/7) of oncofertility websites (p = 0.02). A stand-alone page on oncofertility was present on 28% of oncology websites (19/69), compared with 26% (7/27) of fertility, and 100% (7/7) of oncofertility websites (p < 0.001). In total, 32% (33/103) of websites had this stand-alone page. When comparing between website affiliation categories, no significant differences were found (Table 1). A list with all websites included for qualification can be found in the supplementary on-line material eTable 2^{Table 2}.

Table 2. Websites content: availability of specific oncofertility topics^a.

Items	Oncology (n = 19)	Fertility (n = 7)	Oncofertility (n = 7)	p Value	Professional (n = 10)	Patient (advocacy) (n = 21)	Governmental (n = 2)	p Value
6. Fertility (preservation) options (max 11)^b	7 (5–8)	4 (3–9)	9 (8–10)	0.07	8 (5–10)	7 (5–9)	6 (4;7)^c	0.63
7. Referral information to local clinic(s) or health care professional	11 (58)	4 (57)	5 (71)	0.89	6 (60)	12 (57)	2 (100)	0.73
8. Decision aid tool available	3 (16)	2 (29)	4 (57)	0.12	4 (40)	5 (24)	0 (0)	0.44
9. Moral and ethical considerations	10 (53)	3 (43)	5 (71)	0.71	6 (60)	12 (57)	0 (0)	0.39
10. Risks and benefits	14 (74)	5 (71)	7 (100)	0.50	8 (80)	17 (81)	1 (50)	0.49
11. Success rates	11 (58)	5 (71)	6 (86)	0.48	8 (80)	13 (62)	1 (50)	0.49
12. Costs of reimbursement facilities	4 (21)	2 (29)	4 (57)	0.29	3 (30)	7 (33)	0 (0)	>0.99
13. Audiovisual resources	3 (16)	4 (57)	4 (57)	0.03	6 (60)	5 (24)	0 (0)	0.11
14a. Social media links	19 (100)	6 (86)	7 (100)	0.42	9 (90)	21 (100)	2 (100)	0.36
14b. How many of 6 social media links does the website have?^b	4 (3–5)	4 (1–5)	3 (2–4)	0.49	3 (2–3)	4 (3–5)	5 (6;4)^c	0.04^d
15. HONcode certification	4 (21)	1 (14)	0 (0)	0.80	1 (10)	4 (19)	0 (0)	>0.99

Key.

^a Data presented as Number (%). Websites divided by oncology, fertility and oncofertility organizations as well as affiliation;.

^b Data presented as Median (interquartile range, IQR);.

^c IQR not calculated due to small sample, exact numbers used;.

^d No statistically significant comparisons using Bonferroni correction.

Website content: availability of specific oncofertility topics

All 33 websites that provided a stand-alone page on oncofertility were further appraised for specific topics covered (Table 2). When identifying fertility (preservation) options, most options were mentioned by oncofertility websites with a median of 9 [IQR, 8–10], whereas a median of 7 [IQR, 5–8] were mentioned by oncology websites, and a median of 4 [IQR, 3–9] were mentioned by fertility websites (p = 0.07). The highest median was found among professional organizations (8 [IQR, 5–10]), followed by advocacy organizations (7 [IQR, 5–9]), then governmental organizations (6 [4;7]) (p = 0.63).

Of a total of 11 possible options, those most often mentioned were embryo freezing, followed by ovarian tissue freezing, then egg freezing (see supplementary on-line material, eTable 3^{Table 3}). Options least mentioned were ovarian shielding and radical trachelectomy. Topics most covered were risks and benefits of fertility preservation and success rates with no significant difference between website categories (Table 2). Least available were decision aid tools, information regarding costs or reimbursement facilities, and audiovisual resources on fertility preservation. When comparing the number of specific topics discussed out of six possible categories (referral information, decision-aid tools, moral and ethical considerations, risks and benefits, success rates, costs), a median of 2 [IQR, 1–4] were mentioned by oncology websites, 3 [IQR, 2–5] by fertility websites and 4 [IQR, 4–6] by oncofertility websites (p = 0.11) (Table 3).

Table 3. Availability of specific topics I7–12^a.

Items	Oncology (n = 19)	Fertility (n = 7)	Oncofertility (n = 7)	p Value	Professional (n = 10)	Patient (n = 21)	Governmental (n = 2)	p Value
Specific topics I 7–12^b	2 (1–4)	3 (2–5)	4 (4–6)	0.11	4 (2–5)	3 (2–5)	2 (1;3)^c	0.51

Key.

^a Data presented as median (interquartile range, IQR). Websites divided by oncology, fertility and oncofertility organizations as well as affiliation.

^b Six topics: referral information, decision-aid tool, moral and ethical considerations, risks and benefits, success rates, costs.

^c IQR not calculated due to small sample, exact numbers used.

Quality of information on oncofertility

When using the DISCERN instrument, comparing reliability among oncology, fertility and oncofertility websites, as well as between website affiliations, medians ranged from 27 to 33 (out of 40), indicating moderate scores between 68% and 83% (Table 4). For quality of information on treatment choices, median scores were 26 [IQR, 20–32] out of a maximum of 35 for oncology websites (74%), 26 [IQR, 21–28] (74%) for fertility-, and highest for oncofertility websites with a median of 30 [IQR, 29–33] (86%) (p = 0.05), respectively. All scores were between 50 and 90%, which indicated moderate scores. When assessing the overall rating of the publication, median scores were high for oncofertility websites (4 [4–5]) and moderate for fertility websites (3 [3–4]) and oncology websites (3 [3–4] (p = 0.06). For the total DISCERN score, oncofertility websites had a median of 69 [IQR, 62–69] (86%), followed by a median of 61 [IQR, 57–67] (76%) for fertility- and 58 [IQR, 53–65] (73%) for oncology websites, indicating all moderate scores (p = 0.09). Usability scores were all moderate (medians between 42 (78%) and 48 (89%). Reliability scores (LIDA) were all between 16 (59%) and 19 (70%), indicating moderate scores. The median readability (FKGL) was 10 [IQR, 10–12] for oncology-, 12 [IQR, 9–12] for fertility- and 12 [IQR, 12–12] for oncofertility websites (p = 0.04). No significant differences were found for quality between professional, advocacy, and governmental websites. All organizations with websites having stand-alone pages on fertility preservation in the cancer setting and its qualifications can be found in the Supplementary material eTables 5 and 6.

Table 4. Quality of web content^a.

Items	Oncology (n = 19)	Fertility (n = 7)	Oncofertility (n = 7)	p Value	Professional (n = 10)	Patient (advocacy) (n = 21)	Governmental^b (n = 2)	p Value
Reliability (DISCERN) (8–40)	30 (28–34)	33 (32–36)	33 (30–36)	0.13	33 (31–35)	31 (28–36)	27 (26;28)	0.15
Quality of information on treatment choices (DISCERN) (7–35)	26 (20–32)	26 (21–28)	30 (29–33)	0.05	29 (25–32)	27 (23–31)	25 (17;33)	0.83
Overall rating of the publication (DISCERN) (1–5)	3 (3–4)	3 (3–4)	4 (4–5)	0.06	4 (3–5)	4 (3–5)	3 (3;3)	0.43
Total score (DISCERN) (16–80)	58 (53–65)	61 (57–67)	69 (62–69)	0.09	62 (59–70)	60 (54–69)	55 (46;64)	0.53
Usability (LIDA) (0–54)	48 (46–51)	42 (38–52)	48 (44–53)	0.28	46 (40–51)	48 (46–52)	47 (51;42)	0.79
Reliability (LIDA) (0–27)	17 (14–18)	16 (14–18)	19 (10–20)	0.82	16 (12–20)	17 (14–19)	17 (17;16)	0.99
Readability (FKGL) (range 3–12)	10 (10–12)	12 (9–12)	12 (12–12)	0.04	12 (9–12)	11 (10–12)	10 (8;12)	0.72

Key.

^a Data presented as median (interquartile ranges). Websites divided by oncology, fertility and oncofertility organizations as well as affiliation.

^b IQR not calculated due to small sample, exact numbers used.

Discussion

In this study, we aimed to assess the availability of English written oncofertility information online and its quality, on websites of oncology, fertility and oncofertility organizations, globally. Patients want their fertility addressed sensitively at the time of diagnosis and backed up by access to written or web-based patient information or a fertility specialist (Muller et al., 2017; Wilkes, Coulson, Crosland, Rubin, & Stewart, 2010). Several studies claim that websites can improve knowledge and help in decision-making (Balthazar et al., 2012; Daniluk & Koert, 2015; Ehrbar et al., 2018; Garvelink et al., 2013; Hershberger et al., 2013; Jona & Gerber, 2010; Lewis, 1999). Unfortunately, fewer than half of oncologists refer patients of childbearing age for fertility preservation (Quinn et al., 2009), and few (13%) provide their patients with educational materials on fertility preservation (Quinn et al., 2012). Since there is an information gap on oncofertility information, and options for preservation are not always discussed timely since the focus is likely on the cancer treatment itself, the internet might be the only source of information regarding a patients’ fertility preservation options. Generally, patients view the internet as an additional source of information that may help them better understand their doctors’ recommendations and advice. Furthermore, it has been found that seeking health information online may actually improve the patient–physician relationship (Tan & Goonawardene, 2017). To address these patient needs, we assessed the current state of Internet resources using validated tools.

Surprisingly, we found 68% of all investigated websites mentioned the effects of cancer (treatment) on fertility, and half of all websites mentioned the option of fertility preservation, with a significant difference among oncology, fertility and oncofertility websites. Furthermore, only a third of all websites provided a specific page on oncofertility, with a significant difference between website orientation: 28% of oncology-, 26% of fertility- and, as expected, 100% of oncofertility organizations. Oncology and fertility-facing websites do not necessarily need to build out their own content, but it is vital they point patients toward high-quality oncofertility resources discussed on specialty sites. Of a total of 11 fertility (preservation) options, those most often mentioned were embryo freezing, followed by ovarian tissue freezing, then egg freezing. The range of fertility sparing options could be improved on websites, but may differ around the globe (Rashedi et al., 2018a, 2018b).

Women especially value information about pros and cons of fertility preservation, moral and ethical considerations, medical procedures, risks, benefits, success rates and costs (Hershberger et al., 2013). In our study, these topics, as well as referral information and decision-aid tools, were infrequently mentioned. In a recent guideline on fertility preservation from the British Fertility Society (Yasmin et al., 2018), the role of psychological and decision support was highlighted, and decision-making aids were recommended. As shown in our study as well as in previous research (Mahmoodi, Bekker, King, Hughes, & Jones, 2016), decision-aid tools are often lacking. That said, new ones are being developed and a list of these existing tools can be found in Table 5 of the Supplementary online material. Thus, one of our conclusions is that more decision-aid tools need to be made publicly available (Jones et al., 2017).

When analyzing the quality of websites, we found moderate scores for reliability and usability. Since oncofertility is an emerging topic (de Roo et al., 2017; De Vos, Smitz, & Woodruff, 2014; Jeruss & Woodruff, 2009; Woodruff, 2010, 2013), this may simply reflect the initial online material can now be updated with more contemporary information. Quality may also be improved by updating the content at an appropriate time interval and citing original sources. Information on treatment choices could be improved by describing each treatment, its benefits and risks, effects on quality of life and support for shared decision-making.

None of the median readability scores had the recommended readability level of 7th–8th grade. This shows that information can be hard to understand, which is consistent with other studies (Merrick, Wright, Pacey, & Eiser, 2012; Rosenberg et al., 2016; Storino et al., 2016; Vargas, Chuang, Ganor, & Lee, 2014), which could be improved by using shorter sentences and simpler vocabulary.

There are several limitations in this study. Since this research aims to analyze the content and quality of global information resources, we assessed websites of (inter)national organizations and, therefore, excluded websites run by local organizations, such as hospitals and clinics, which could be of high-quality. Since we qualified information in the English language, it is possible that websites originally written in another language, but qualified using the English version, might be of lesser quality than in their original language. Also, for further qualification with specific content questions and quality instruments, we only considered websites providing a stand-alone page on oncofertility, so we may have missed some organizations without a stand-alone page, but with links to other organizations which might have high-quality information. Another limitation is that the readability test does not consider medical jargon. An option for websites could be to use an API to link to Repropedia, a reproductive lexicon which defines reproductive health terms commonly used in oncofertility treatment (‘Repropedia: A Reproductive Lexicon’). Despite these limitations, this is the first study to our knowledge to assess the availability and quality of global online resources on female oncofertility.

We suggest future research on oncofertility web content for pediatric patients. In developing literature for patients, it can be challenging to consider different needs at different time points in the cancer treatment process (start of treatment, during treatment and following treatment), as well as cultural, religious and gender or sex differences. Other aspects of care, such as sexual dysfunction and fertility-related psychological distress, might be valuable to include, although not objectively tested here.

In conclusion, the results of this study show that availability of web content on fertility preservation for young women facing cancer is highly variable, reliability is moderate and the recommended readability level is not reached. Although traditional in-person instruction by medical professionals is of huge importance, in this digital age, it is vital that online health information is of high-quality (Tan & Goonawardene, 2017). Since it has been shown that women prefer additional online resources as a part of their health decision-making processes (Muller et al., 2017), physicians could help women identify high-quality information on websites for their own organizations, as well as by recommending additional websites to their patients. Finally, our data support additional work by all stakeholders in the development of new or adapted information. Since the field is still developing, materials that have been created by various partner groups, like the Global Oncofertility Network (Ataman et al., 2016; ‘Global Oncofertility Network’), could be translated or changed to local or national context, shared with emerging groups and refined according to the variables outlined above.

Acknowledgements

The authors would like to thank Patricia Smith, a Northwestern University librarian, for helping with initial web searches to locate sites that qualified for inclusion in the study. The authors also thank high school students Vani Addepalli (VA) and Semira Allen (SA) for their help with some of the web search activities during a summer internship in the Woodruff Lab. They would also like to thank Joanne in ‘t Hout, a Radboudumc statistician, for helping with statistical analysis.

Disclosure statement

No potential conflict of interest was reported by the authors.

Correction Statement

This article has been republished with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by the Center for Reproductive Health After Disease [P50HD076188] from the National Institutes of Health National Center for Translational Research in Reproduction and Infertility (NCTRI).