Access the full text.
Sign up today, get DeepDyve free for 14 days.
Loading next page...
/lp/jmir-publications/knowledge-acquisition-and-social-support-in-online-health-communities-Y86yPagMU4
References
References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.
- Publisher
- JMIR Publications
- Copyright
- Copyright © The Author(s). Licensed under Creative Commons Attribution cc-by 4.0
- ISSN
- 2369-1999
- DOI
- 10.2196/39643
- Publisher site
- See Article on Publisher Site
Abstract
Background: Patients and caregivers widely use online health communities (OHCs) to acquire knowledge from peers. Questions posed in OHCs reflect participants’ learning objectives and differ in their level of cognitive complexity. However, little is known about the topics and levels of participants’ learning objectives and the corresponding support they receive from members of OHCs. Objective: This study aimed to investigate the knowledge acquisition of patients and caregivers in an OHC. Specifically, we investigated the distribution and topics of posts with learning objectives at different cognitive complexity levels, the type and amount of social support provided to meet users’ learning objectives at different cognitive complexity levels, and the influence of social support on the change in learning objectives. Methods: We collected 10 years of discussion threads from one of the most active ovarian cancer (OvCa) OHCs. A mixed methods approach was used, including qualitative content analysis and quantitative statistical analysis. Initial posts with questions were manually classified into 1 of the 3 learning objectives with increasing cognitive complexity levels, from low to high, based on the Anderson and Krathwohl taxonomy: understand, analyze, and evaluate. Manual content analysis and automatic classification models were used to identify the types of social support in the comments, including emotional support and 5 types of informational support: advice, referral, act, personal experience, and opinion. Results: The original data set contained 909 initial posts and 14,816 comments, and the final data set for the analysis contained 560 posts with questions and 3998 comments. Our results showed that patients with OvCa and their caregivers mainly used OHCs to acquire knowledge for low- to medium-level learning objectives. Of the questions, 82.3% (461/560) were either understand- or analyze-level questions, in which users were seeking to learn basic facts and medical concepts or draw connections among different situations and conditions. Only 17.7% (99/560) of the questions were at the evaluate level, in which users asked other OHC members to help them make decisions or judgments. Notably, OvCa treatment was the most popular topic of interest among all the questions, regardless of the level of learning objectives. Regarding the social support received for different levels of learning objectives, significant differences were found in the advice (F =9.69; P<.001), opinion (F =11.56; P<.001), and emotional 2437.84 2418.18 support (F =3.24; P=.01), as determined by one-way ANOVA, whereby questions at the evaluate level were more likely to 2395.88 receive advice, opinion, and emotional support than questions at the lower levels. Additionally, receiving social support tends to drive users to increase the cognitive complexity of the learning objective in the next post. Conclusions: Our study establishes that OHCs are promising resources for acquiring knowledge of OvCa. Our findings have implications for designing better OHCs that serve the growing OvCa community. https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 1 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al (JMIR Cancer 2022;8(3):e39643) doi: 10.2196/39643 KEYWORDS online health community; ovarian cancer; health information needs; social support; knowledge acquisition high (ie, cognitively simple to complex), the 6 levels are Introduction remember, understand, apply, analyze, evaluate, and create. The theory assumes that to achieve a higher level of learning Background objectives, one must master the lower levels. Online health communities (OHCs), also known as online This study chose 3 levels, understand, analyze, and evaluate, support groups, are 1 of the 3 primary channels for health rather than adopting all 6 levels because they are close to real consumers seeking health information on the web in addition web-based health information–seeking scenarios. As found in to search engines and health professionals [1]. Numerous studies the analysis by Cartright et al [10], of queries from web search have provided substantial evidence that patients benefit from engines, there are 2 representative web-based health OHC participation [2-5]. OHCs facilitate information exchange information–seeking intentions: evidence based and hypothesis and knowledge acquisition among users. For people with cancer directed. With the evidence-based intention, one mainly focuses and their caregivers, who have a constant and evolving need on locating information regarding signs and symptoms, which for information, OHCs are particularly important for can be mapped to the understand level of learning. The around-the-clock availability, immediate and asynchronous hypothesis-directed intention, which drives individuals to draw communication, and anonymity [6,7]. connections and discriminate among different uncertain Users ask questions on OHCs for knowledge acquisition. situations and conditions, aligns with the analyze level. Finally, Questions posed by patients to acquire knowledge to meet their the evaluate level corresponds to the decision-making intention, learning objectives vary in cognitive complexity. The cognitive which involves seeking information to make a treatment complexity of learning objectives describes the cognitive skills decision. and abilities the learner desires to achieve. For example, a Reciprocity is another substantial benefit of OHCs [11,12]. question seeking advice on treatment decisions from peers (eg, Knowledge building and collaborative knowledge production surgery vs biological therapies) is cognitively more complex take place through discourse among members of OHCs [13]. than one looking for facts in medical directions (eg, how many Peer users of the community, who usually face the same health times a day is a pill to be taken). To identify the cognitive condition and endure a similar experience, can provide social complexity level of learning objectives in OHC users’ questions, support by replying to the initial questions and follow-up this study borrowed the Anderson and Krathwohl taxonomy of discourse [3,14]. We focus on the 2 most frequently exchanged learning (A&K taxonomy) [8] from educational psychology. types of social support in OHCs: informational support (ie, This taxonomy was first proposed by Bloom in 1956 [9] and offers information, such as the course of the condition, later revised by Anderson and Krathwohl [8]. As shown in treatment, finance, and insurance) and emotional support (ie, Figure 1, the A&K taxonomy defines 6 levels of learning expresses emotions such as caring and concern) [5,6,15]. objectives with increasing cognitive complexity. From low to Figure 1. Adapted from the Anderson and Krathwohl taxonomy of learning [8]. to achieve cognitively complex or simple learning objectives. Objectives Topics and health conditions discussed in OHCs may affect the Because OHCs are a promising learning resource for patients patients’ learning objectives. Savolainen [16] found that >70% and caregivers, an in-depth study of users’ learning objectives of the questions in OHCs for depression sought an opinion or and the corresponding support they receive is needed. First, it evaluation of an issue, resembling a high-level learning must be examined whether patients and caregivers use OHCs objective, whereas contrasting results were found in an OHC https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 2 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al for alcoholism, where approximately 50% of the posts looked However, owing to a lack of disease awareness, 69% of the for factual information that serves low-level learning objectives patients with OvCa had not heard of or knew nothing about [17]. However, there is scarce literature regarding the learning OvCa before their diagnosis, thus making the knowledge objectives of users of OHCs for cancers. To deal with the acquisition and learning process extremely difficult [26]. numerous physical and psychosocial consequences of survival, Furthermore, studies of people living with OvCa are relatively patients with cancer and their caregivers have been using OHCs limited, although people with OvCa need a lot of support. There to address various cancer-related information needs and gain is a dearth of research investigating what information individuals knowledge about cancer [18-20]. An examination of the learning with OvCa who use OHCs wish to acquire and what support objectives of people with cancer will add to the empirical they receive. The findings of this study also contribute to the knowledge on how OHCs facilitate knowledge acquisition for knowledge on how to better support the OvCa community. patients with different health conditions. Methods Second, it is unclear whether all levels of learning objectives are well supported in OHCs. Higher levels of learning objectives Research Setting: National Ovarian Cancer Coalition (eg, evaluate) are more difficult to achieve than lower levels of CancerConnect Community learning objectives (eg, understand) and require support from We collected data from CancerConnect, an OHC for patients skilled and knowledgeable peers [17,21]. In this study, we with OvCa, managed by the National Ovarian Cancer Coalition examined the type and amount of support for different levels (NOCC). NOCC is a nonprofit OvCa advocacy organization of learning objectives by measuring the corresponding social that has devoted itself to educating and supporting patients with support qualitatively and quantitatively. OvCa, survivors, and caregivers since its inception in 1991. The Third, we are interested in investigating how users’ learning NOCC CancerConnect Community is one of the most active objectives change during their participation in an OHC. OvCa OHCs [27]. It is a peer-supported OHC with the goal of Moreover, if one’s learning objective is well supported by peers providing an open-access platform that encourages and enhances in the OHC, will this drive them to modify their learning interpersonal learning via informational and emotional peer objective to ask a more cognitively complex question in the interactions. To this end, NOCC allows registered users to OHC? The answers to these questions will shed light on the participate and contribute to the community in several ways, effectiveness of OHCs and the designing of OHCs as web-based such as initiating and replying to posts, searching and reading learning resources. posts and comments, creating profiles, joining groups, and sending and receiving private messages. Therefore, this paper seeks to answer the following research questions (RQs): Ethical Considerations • RQ1: What are the distributions and topics of posts at This study was reviewed and approved by the Institutional different levels of learning objectives? Research Board of University of Pittsburgh (STUDY20040102). • RQ2: What type and amount of social support are provided In addition, permission was obtained from NOCC to conduct to posts at different levels of learning objectives? this study. • RQ3: How do users’ learning objectives change during their Data Analysis participation in an OHC? Is the change in the learning Our NOCC data set contained 909 OvCa discussion threads objectives of users associated with the type and amount of posted between June 2010 and December 2020. Each thread social support received? was made of 1 initial post and corresponding comments if any. To answer these RQs, we collected 10 years of discussion In total, there were 909 initial posts and 14,816 comments. threads from an OHC for patients with ovarian cancer (OvCa) Figure 2 illustrates the overall data analysis process. We first and caregivers. Because OvCa is a rare cancer [22], health performed manual annotations on the 909 initial posts to information seeking on the internet can be particularly determine whether there was a question articulated in the post. challenging because of information scarcity and limited public As a result, 560 posts and their 3998 comments were retained awareness. In addition, OvCa is the deadliest cancer among for further analysis. The posts without any questions mainly women [22]. The 5-year relative survival rate of patients with consisted of sharing personal updates, sharing resources, OvCa from 2011 to 2017 in the United States was 49.1% [23]. provoking discussions, and providing inspiration. The posts For individuals with OvCa and their families, managing this were then coded in terms of the level of the learning objective cancer can be stressful because of intensive treatments and high and OvCa-related topics. For the 3998 comments on the initial rates of disease progression [24]. Owing to limitations in early posts, we first performed manual annotation on 500 randomly detection, OvCa is often diagnosed at late stages when the chosen comments to identify the types of social support. likelihood of cure is low. In the United States, it is the most Automatic classification models were then trained and applied common cause of death due to gynecological malignancies [25]. to predict different types of social support in the remaining People with OvCa use OHCs to address their OvCa-specific, comments. treatment-related, and coping-related information needs [19]. https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 3 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al Figure 2. Data analysis process. OvCa: ovarian cancer. The real username and user profile image are removed for Identifying the Level of Learning Objective privacy. As mentioned earlier, we borrowed 3 levels from the A&K Two coders (YC and KT) applied the coding framework to 100 taxonomy of learning [8] to identify the level of learning sample posts to determine the level of the learning objective objectives in the users’ questions. The descriptions of each level that best describes the cognitive complexity of the questions. of the learning objective and the deidentified example questions Substantial agreement was achieved between the 2 coders on are displayed in Table 1. To achieve higher levels in the A&K the 100 sample posts (percentage agreement=0.79; Cohen taxonomy, one must master the lower levels in the hierarchy. κ=0.72), indicating an acceptable level of agreement [28,29]. Therefore, the 3 levels of learning objectives were coded The 2 coders then met to discuss any disagreements. Throughout mutually exclusively. For example, Figure 3 shows a post with the discussion, all disagreements were addressed, and no the evaluate level of learning objective, as the poster described changes were made to the codebook. A coder annotated the her situation and sought decision-related information from peers. remaining posts by using the codebook. Table 1. Coding framework of learning objective in the initial post. Learning objective Description Example question Understand Pursuit of facts, concepts, and ideas by describing, “Hi does anyone have information on AMG 386? Thank You” explaining, identifying, detailing, interpreting, summarizing, and so on Analyze Pursuit of connections and relationships among “I recently developed small red dots all over my legs, look like little blood multiple concepts by differentiating, comparing, marks. I’m on Avastin and wonder if anyone has experienced these marks distinguishing, contrasting, sorting, and so on on their body?” Evaluate Pursuit of decision or judgment given specific “Hi Sisters, I finished front line 12/8, and ca has be tested 3 times since. conditions by appraising, arguing, judging, select- The last one showed 2 point increase and Dr wasn’t concerned as said ing, critiquing, weighing, recommending, assessing, basically save number 28 to 30. This was 1/22. Today it has went up .8. predicting, and so on Any reason to be concerned since trend is upward? I’m concerned of this continuing and I’m already full of worry.” https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 4 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al Figure 3. An example of an evaluate-level question. Using this framework, the 2 coders individually annotated all Identifying the OvCa Topics of Questions the posts. Questions in each post included 1 or multiple topics. To better understand OvCa users’ information needs at different Later, topics that appeared in <10 posts were further grouped levels of learning objectives, the topics of the questions in the into Others. Consequently, 9 codes were used to classify the initial posts were annotated through content analysis. The coding topics of information needs in the initial posts (Table 2). An framework was inductively developed by a nurse practitioner acceptable interrater agreement was obtained between the 2 by immersing herself in the posts. A coding framework with coders, with an average percentage agreement of 0.94 and Cohen 13 topics was established initially. κ coefficient of 0.72, ranging from 0.62 to 0.81 across 9 categories [28,29]. The 2 coders discussed and resolved all disagreements and reached an agreement in all cases. Table 2. Coding framework of topics of questions. Topic Description Code Disease management Information needs related to ovarian cancer disease management, such as diagnosis, prognosis, finding gyneco- DM logic oncologist, preparing for visit, advance care planning or advance directives, borderline malignant tumors, prophylactic surgery, secondary prevention, monitoring for recurrence, management of recurrence, and supportive care or palliative care Symptom management Information needs related to ovarian cancer symptom management, such as fatigue, sleep, bowel, pain, neuropathy, SM cognitive memory, nausea, vomiting, bloating, ascites, appetite, appearance, shortness of breath, lymphedema, urinary, early menopause, ostomy management, rash, anemia, mouth sore, and myelosuppression Treatment Information needs related to ovarian cancer treatment, such as medications, surgery, radiation, chemotherapy, TM biological therapies, and clinical trials Treatment decision Information needs related to ovarian cancer decision-making, such as how to make treatment decisions TD Emotional management Information needs related to emotional management, such as anxiety, depression, fear of recurrence, mood EM swings, coping, grief, and loss Self-management Information needs related to self-management, such as nutrition, spiritual support, physical activity, and rela- SF tionship with loved ones Practical needs Information needs related to practical needs, such as finance, insurance, employment, legal, and community PN resources Caregiving Information needs related to caregivers’ needs, such as stress, caregiver coping, grief, and loss CG Others Other ovarian cancer–related information needs, such as communication, sexuality, rehabilitation, complementary OT therapy and integrative medicine, ovarian cancer organization, and facilities DM: disease management. SM: symptom management. TM: treatment. TD: treatment decision. EM: emotional management. SF: self-management. PN: practical needs. CG: caregiving. OT: others. https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 5 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al fact and “0” for all other types of informational and emotional Identifying the Types of Social Support support. The second comment was coded as “1” for providing The 2 most common types of social support exchanged in OHC a fact and an advice and “0” for all others. are informational and emotional support [5,6]. In this study, as the aim was to investigate what information users receive as The social support types provided in the 3998 comments were answers to their questions, the informational support provided identified in 3 steps. First, 2 coders coded 150 sample comments in the comment was further classified by using the framework to ensure the reliability of the coding framework. On average, proposed by Chuang and Yang [17]. Chuang and Yang [17] an agreement rate with percentage agreement of 0.94 and Cohen identified five types of informational support: κ of 0.84 were achieved, indicating an almost perfect agreement [28,29]. Second, after addressing all disagreements, a coder Advice: the comment offers ideas, suggestions, and actions coded 350 more comments. As a result, a data set of 500 to cope with challenges. comments was obtained, in which each comment contained a Referral: the comment refers to information sources such comment text and corresponding support labels. Third, as it as books, websites, and contacts. would be impractical to annotate all 3998 comments, the Fact: the comment offers facts or reassesses the situation. decision was made to build machine learning–based classifiers Personal experience: the comment shares personal stories by using the already annotated comments. In total, 6 machine or incidents. classifiers were built for each support type. A pretrained Opinion: the comment offers a view or judgment about Bidirectional Encoder Representations from Transformers something. However, this is not necessarily based on facts (BERT) language model [30] was fine-tuned for each or knowledge. classification task. BERT was used because it obtained good In addition, emotional support was marked if a comment classification accuracy with less data on different downstream provided empathy, encouragement, or appreciation [12]. text classification tasks, such as sentiment and emotion classification [30]. The data set was split into 3 folds with a All 6 types of social support, including emotional support and 70:10:20 ratio for training, validation, and testing, respectively. 5 types of informational support, were coded in a binary fashion, The accuracy reported in Table 3 is based on the testing fold. and a comment could provide 0, 1, or multiple types of support. The interrater agreement between the 2 coders and performance If no informational or emotional support could be identified, of the classification models are presented in Table 3. The code the comment was coded as “Others.” For example, Figure 4 for the model and access to our model are listed on GitHub [31]. displays 2 comment examples that replied to posts shown in Finally, the models were applied to predict the social support Figure 3. The first comment was coded as “1” for providing a types for the remaining comments. Figure 4. Examples of comments. https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 6 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al Table 3. Interrater agreement between human annotators and classification score for social support types in the comments. Support type Interrater agreement Support type prediction Percentage agreement Cohen κ Precision Recall F-score Advice 0.96 0.88 0.77 0.85 0.81 Referral 0.98 0.94 0.82 1.00 0.90 Fact 0.93 0.86 0.82 0.77 0.79 Personal experience 0.90 0.80 0.95 0.87 0.91 Opinion 0.93 0.79 0.81 0.81 0.81 Emotional support 0.91 0.82 0.91 0.74 0.82 Others 0.95 0.76 N/A N/A N/A Average 0.94 0.84 0.85 0.84 0.84 N/A: not applicable. Number of Topics Results In most of the initial posts, users tended to seek information Overview and knowledge about 1 (363/560, 64.8%) or 2 (176/560, 31.4%) topics per post. There were only 21 posts in which users Of 909 initial posts, 560 (61.6%) were associated with learning consulted their peers on >2 OvCa topics (21/560, 3.8%). objectives, as indicated by the questions asked in the posts. The following results were based on the analysis of the 560 initial The initial posts were grouped according to the 3 levels of posts with identified learning objectives and 3642 comments learning objectives; the average number of topics in each group that provided at least one type of support. is presented in Table 4. A one-way between-subject ANOVA was performed on the number of topics in 1 post as a function Learning Objectives in the Initial Posts (RQ1) of the level of learning objective. With violation of the Distribution of Users’ Learning Objectives in the Initial assumption of homogeneity of variance, an F-test with Brown-Forsythe adjustment was conducted. The results Posts suggested a statistically significant difference in the number of Among the 560 posts with questions, the analyze objective was topics among the different levels of learning objectives the most common, accounting for almost half of the total (F =72.54; P<.001). A Games-Howell post hoc test 2193.364 (257/560, 45.9%). Following this, 36.4% (204/560) of the posts revealed that there were significantly more topics in the posts with questions sought understand-level knowledge, whereas asking for an evaluate-level learning question (N=1.83; P<.001) evaluate, the most complex learning objective, only accounted than in posts with the analyze-level learning objective (N=1.50; for 17.7% (99/560) of the question-asking posts. This result P<.001). The posts seeking understand-level knowledge suggests that people with OvCa mainly use the NOCC consisted of the least number of topics compared with the 2 community to look for simple knowledge, such as facts, higher levels (N=1.05; P<.001). The difference in the number concepts, or relationships between facts and concepts, rather of topics may indicate that people with OvCa tend to acquire than complex knowledge relating to treatment decisions and information across multiple topics to obtain evaluate-level judgments. knowledge. By contrast, for lower-level learning objectives, their information needs were more likely to focus on 1 specific topic. Table 4. Number of topics per post at each level of learning objective. Learning objective Topics per post, mean (SD) Posts, n (%) Understand 1.05 (0.24) 204 (36.4) Analyze 1.50 (0.54) 257 (45.9) Evaluate 1.83 (0.73) 99 (17.7) Total 1.40 (0.57) 560 (100) levels of learning objectives to show what OvCa-related Category of Topics knowledge patients and caregivers wanted to acquire. Then, for Using the coding framework in Table 2, the questions in the posts with >1 topic, the frequencies of all topic pairs were initial posts were classified into 9 categories based on examined to further demonstrate what topics tended to be OvCa-related topics. In this section, 2 results for the topic inquired about together. categories are presented. First, topics were grouped by different https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 7 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al Figure 5 shows the distribution of the 9 OvCa-related topics at treatment decisions accounted for a significant portion (30/181, each level of learning objective. Each bar represents the posts 16.6%) of the evaluate level. However, it is questionable of 1 of the 3 levels of learning objectives, whereas segments in whether users should use OHC as a resource for making the bar denote the portion of a topic among all posts with the treatment-related decisions. Emotional management and same level of learning objective. Segments of the same color practical needs presented similar patterns: the proportions of were comparable. understand and evaluate questions were higher than that of analyze questions. Caregiving information accounted for a much It is evident that treatment is the most popular topic of interest greater share of understand questions than the other two. Finally, in all knowledge acquisition posts, with a higher proportion in the ratios of the other topics were similar for all 3 levels of the analyze level (175/385, 45.4%) than in the other 2 levels of learning objectives. learning objectives. This result indicated that comparing or differentiating treatment information was a common need among Chi-square results revealed a significant association between people with OvCa in OHCs. In addition, pursuing treatment the levels of learning objectives and the topics of disease information to understand or evaluate was frequent, which might management (χ =17.2; P<.001), symptom management be because the treatment information of OvCa was complex 2 2 (χ =40.2; P<.001), treatment (χ =38.6; P<.001), treatment 2 2 and scattered, making the topic of treatment the dominant decision (χ =85.8; P<.001), and emotional management information needed across all the learning objectives. Analyzing symptom management is the second most prevalent information (χ =7.7; P=.02). However, no significant association was found needed, whereas understanding and evaluating symptom between the learning objective levels and topics of management information is not that popular. The results suggest 2 2 self-management (χ =0.0; P=.99), practical needs (χ =0.3; 2 2 that for symptom management, patients and caregivers struggle 2 2 P=.19), caregiving (χ =0.4; P=.09), and others (χ =0.6; more with the differentiation or connection among different 2 2 symptoms than with learning about basic symptoms or making P=.71). judgments. Figure 6 shows the proportions of different topic pairs among On the contrary, disease management was more associated with the 245 topic pairs extracted from questions with >1 topic. the understand and evaluate levels of learning objectives than Notably, treatment and symptom management were most likely the analyze level, implying that people with OvCa needed to appear together in a single post (72/245, 29.4%). In addition, support for interpreting disease information such as diagnosis, patients with OvCa and their caregivers tended to learn about prognosis, and recurrence on both a basic fact or concept level treatment along with disease management or treatment decisions. and a higher judgment or decision level. It is notable that Figure 5. Distribution of ovarian cancer topics at each learning objective level. A: analyze; CG: caregiving; DM: disease management; E: evaluate; EM: emotional management; OT: others; PN: practical needs; SF: self-management; SM: symptom management; TD: treatment decision; TM: treatment; U: understand. https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 8 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al Figure 6. Co-occurrence of topic pairs in 1 post (darker color indicates larger proportions). CG: caregiving; DM: disease management; EM: emotional management; OT: others; PN: practical needs; SF: self-management; SM: symptom management; TD: treatment decision; TM: treatment. belonging to each learning objective. Log transformation is Social Support in the Comments (RQ2) applied to the total number of each type of comment and plotted in the line chart. In general, the largest number of supportive Number of Replies to Posts at Different Levels of replies was provided to posts with the evaluate-level learning Learning Objectives objective, followed by the understand level, and it was the least The 3642 comments providing support were grouped based on for the analyze-level learning objective. the learning objective in the post. Posts with the understand As determined by one-way ANOVA, significant differences level were likely to receive the largest average number of among the 3 levels of learning objectives were found in advice comments from peers (N=7.68), followed by the evaluate (F =9.69; P<.001), opinion (F =11.56; P<.001), and (N=7.07) and analyze (N=5.63) levels. However, the results of 2437.84 2418.18 the one-way ANOVA suggested no statistically significant emotional support (F =3.24; P=.01) levels. A 2395.88 difference between the average number of comments among Games-Howell post hoc test revealed that posts seeking the 3 levels of learning objectives (F =2.712; P=.07). 2451.295 analyze-level knowledge received significantly less opinion support compared with understand-level (P=.002) and Social Support Provided for Posts at Different Levels of evaluate-level posts (P<.001). The amount of advice support Learning Objectives at the evaluate level was significantly higher than that at the The types and amount of social support provided by the repliers analyze (P<.001) and understand (P=.001) levels. For in each comment were aggregated by posts. Figure 7 shows the emotionalsupport, a significant result was found only between number of different types of support received in each post analyze and evaluate (P=.02) levels. https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 9 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al Figure 7. Type and amount of social support provided for questions at each learning objective level. likely to increase or remain at the same level of learning Influence of Social Support on Change in the Learning objectives as they continued posting, asking questions, and Objective (RQ3) acquiring knowledge in the same forum. Overview We also examined the specific types of transitions from different Because some users posted >1 posts with learning objectives levels of learning objectives (eg, from understand to in NOCC, this allowed the researcher to unveil how the learning understand). This helped reveal how the current level of learning objectives of the same user change over time. In total, 344 objective affected the subsequent post’s learning objective. distinct users posted 560 posts with learning objectives. Most First, from analyze to analyze (A→A: 57/216, 26.4%) was the users (244/344, 70.9%) posted only 1 post, and 29.1% (100/344) most common transition. The amount and ratio are also higher of users posted multiple posts. Among the 100 users who posted than those from analyze to understand (A→U: 22/216, 10.2%) >1 posts with learning objectives, 60, 17, 9, and 14 posted 2, 3, and analyze to evaluate (A→E: 24/216, 11.2%), suggesting that 4, and >5 posts, respectively, with learning objectives. These analyze-level questions were likely to be followed by another 100 users were further examined to uncover changes in their analyze-level question than the increase or decrease in levels learning objectives in the NOCC and the influence of social of learning objectives of the same user. Second, after asking an support on the change. understand-level question, users tended to increase the level of learning objective and ask an analyze-level question (U→A: The change in the learning objective is defined as the transition 36/216, 16.7%). This possibility is higher than asking another between the level of the learning objective in post P and post understand-level question (U→U: 27/216, 12.5%) or P for the same user U. The change in learning objectives was i+1 evaluate-level question (U→E: 10/216, 4.6%). This might be classified into 3 categories based on the transition from post P i attributed to the fact that the understand-level learning objective to P : knowledge increase, knowledge decrease, and no was relatively easy to achieve, or the users’ OvCa-related i+1 knowledge might evolve and increase over time, driving them change. For example, if a user posted 3 initial posts (ie, P , P , 1 2 to pursue a higher level of learning. Third, evaluate-level posts and P ) in the NOCC forum and the level of learning objective were mainly followed by analyze-level posts (E→A: 22/216, in them are P —understand, P —analyze, and P —analyze, 1 2 3 10.2%) or understand-level posts (E→U: 12/216, 5.6%). Only then the change in learning objective from P to P is knowledge 1 2 rarely would users ask another evaluate-level question (E→E: increase, and the change from P to P is no change. In total, 2 3 6/216, 2.8%). In addition, users were more likely to increase 216 changes in learning objectives were identified from the 100 the learning objective by 1 level (ie, U→A: A→E) or decrease users who contributed multiple posts in the NOCC forum. it by 1 level (ie, E→A: A→U) in 2 consecutive posts than to increase or decrease it by 2 levels (ie, U→E: E→U), indicating Change of Learning Objectives of the Same User that the change in learning objectives was a gradually evolving In general, 41.7% (90/216) of the pairs of 2 consecutive posts process. sought information on the same level of learning objectives, which resulted in no change. Knowledge increase, in which the Social Support Received and Change of Learning learning objective in the subsequent post was higher than the Objective previous one, was the second most frequent (70/216, 32.4%). Figure 8 shows how the type and amount of social support The least frequent type of change was knowledge decrease received for the current post influenced users’ learning (56/216, 25.9%). It can be inferred that NOCC users were more objectives in the next post. On average, for most types of social https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 10 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al support, when users received more support, including advice, rather than decrease or maintain the same level of learning personal experience, opinion, and emotional support, they were objective. No statistically significant differences were found more likely to increase their learning objective in the next post, between the 3 types of changes. Figure 8. Amount and type of social support received and change in the learning objective level. and the information is complex and scattered. This might cause Discussion OHC users to seek basic facts and concepts at the understand level more often. In addition, the prevalence of analyze-level Overview questions could be explained by the fact that OvCa is a complex This study investigated knowledge acquisition by people with disease. Because the diagnosis, staging, and treatment are OvCa in an OHC. We borrowed three levels of learning complex, patients and caregivers have to learn and sort out objectives from the A&K taxonomy: understand, analyze, and which information applies to them and which does not. For evaluate. The results revealed (1) the distributions and topics example, on average, women with OvCa under treatment need of posts at different learning objective levels, (2) the type and to manage 12 concurrent symptoms [32]. amount of corresponding social support at each level, and (3) Regarding OvCa-related topics, treatment is the most popular the influence of social support on changes in learning objectives. topic of interest among all the information needs, regardless of The principal findings, contributions, implications, and the level of learning objectives. This finding is in accordance limitations of this study are discussed in the following sections. with the results in the study by Madathil et al [19], in which Principal Findings treatment-related information was found to be the most Our results showed that NOCC was mainly used by patients sought-after information by patients (41.3%) compared with with OvCa and their caregivers to address information needs OvCa-specific and coping information. Data analyses were with low- to middle-level learning objectives. Of the questions, conducted at the Ovarian Cancer National Alliance, another 82.3% (461/560) were either at the understand or analyze levels OHC for OvCa. We identified 9 different topics by using our of cognitive complexity, in which the user initiates a post to fine-grained topic classification framework, and the posts were pursue basic facts and concepts or connections and relationships classified in a nonmutually exclusive manner. Treatment was among multiple concepts. Notably, only 17.7% (99/560) of the still found to be the most popular topic. This finding further posts with questions were associated with an evaluate-level underlines the high demand for treatment-related information learning objective, in which the users asked other OHC members and support among people with OvCa. It is also noteworthy that to help them make decisions or judgments based on their specific treatment decision accounted for a large share at the evaluate conditions. These results are partially different from the findings level despite the concern that an OHC might not be an in [16], where >70% of the posted questions in the web-based appropriate resource to ask for treatment-related decisions. Such discussion forums sought an opinion or evaluation of an issue, findings add to the demand for research efforts to assess the resembling an evaluate- or analyze-level question, whereas the quality of treatment-related decisions shared by peers in OHCs. need for factual and procedural information was less common. In addition, we examined the type and amount of informational These conflicting results could be attributed to the different support in the comments, providing a means to study the health conditions studied. In in the study by Savolainen [16], quantity and quality of information that OHC users can acquire the topic of interest in the threads was depression, whereas in at different levels of knowledge acquisition. In general, users this study, it was OvCa, which is listed as a type of rare cancer in the NOCC group received the largest number of comments by the National Institutes of Health [1]. Therefore, the general for understand-level learning objective (N=7.68), followed by public lacks disease awareness and education regarding OvCa, https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 11 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al evaluate-level (N=7.07) and analyze-level (N=5.63) learning Implications for OHC objectives. However, the number of comments itself was not First, although there is an extensive body of literature enough to reflect the quality and quantity of social support in investigating OHCs, and it has been proven that patients and the OHC because a comment could provide 0, 1, or multiple their caregivers would use OHCs to post questions and acquire types of social support; therefore, we classified the types of knowledge [12,15,17], little has been done to differentiate social support in the comment, especially informational support. knowledge acquisition with different levels of learning objectives and the associated social support provided by peers Descriptive results indicated that, in general, the total amount in OHCs. Our study contributes empirical evidence and of social support of all types was the largest for evaluate-level demonstrates that user interactions in OHCs can be described learning, followed by understand-level learning, and it was the and studied from a knowledge acquisition perspective. Not all least for analyze-level learning. For each type of social support, information needs regarding the underlying cognitive complexity fact was acquired the most compared with other types of support. of the learning objectives are identical. Our study also This result is consistent with the results in the study by Chuang demonstrated that OHC is a promising resource for users to et al [17], which were based on a manual analysis of an address information needs with different cognitive complexities alcoholism OHC. Regarding the effect of the learning objective, and that OHCs can help users to improve knowledge if their the results suggest that more advice, opinions, and emotional information needs are well supported with informational and support were obtained for questions seeking evaluate-level emotional support from peers. learning. A possible explanation for this finding is that some subjective knowledge, to a certain extent, was needed to support Correspondingly, OHCs ought to recognize the cognitive people with OvCa’s information needs of evaluate-level complexity of the user’s information needs and the underlying learning. As justified by the interviewees in the study by Harkin learning objective. Importantly, the quality and quantity of social et al [2], practical advice shared by peers in OHCs was support from peers are critical for users to address their welcomed by many interviewees, as such information led them information needs and seek higher-level knowledge. Enhancing on a “journey to become informed.” It is also notable that patients’ learning objectives is important because pursuing although the questions with the analyze-level learning objective cognitively more complex learning objectives implies higher were the most frequently posted in the OHC, they received the patient activation—informed and activated patients who actively smallest number of average comments and the least amount of engage in health care and decision-making. Higher patient almost all types of social support in the comments. Measures activation is associated with better health-related outcomes beyond the number of comments and support are required to [34,35]. Given the result that certain types of support were explore this finding in the future. associated with an increase in learning objectives, algorithms or human moderators in OHCs are expected to match the level Finally, we examined multiple posts from the same user, and of learning objectives in the original post with the appropriate the results demonstrated that OvCa users’ learning objectives types of social support from active peers. changed during OHC use. This change was reflected by the transition from the current post’s learning objective to the With their social features, OHCs amplify the benefits of a wealth subsequent post’s learning objective. Most of the users who of information as well as the negative emotions shared by peers. posted >1 post with a learning objective in the NOCC tended In addition, there are concerns about the quality of the narratives to increase their learning objective (70/216, 32.4%) or remained shared by patients in OHCs [36,37]. False information and at the same level of learning objective (90/216, 41.7%), as they rumors can cause false expectations [2]. To deal with the continued posting and seeking information in the same forum. downside of OHCs, it is suggested that the content be carefully Furthermore, for users who increased their learning objective administered by moderators with professional backgrounds. in the next post, a larger amount of support in advice, personal Attention should be devoted to information-seeking posts with experience, opinion, and emotional support was observed in high cognitively complex learning objectives such as pursuing the current post (Figure 8). In other words, receiving more social judgments and decisions from peers. In addition, some support might drive the users to acquire higher-level knowledge high-quality learning materials can be developed and in the same OHC. Although the result was not statistically disseminated via OHCs, as they have been proven to be an significant, this finding adds to previous studies that have active informal learning platform. demonstrated the effect of social support on member retention Implications for OvCa Community and engagement [5,6,33] and contributes new evidence on the potential effects of social support on collaborative knowledge People with OvCa have exhibited constant and dynamic building and generation in web-based communities [13]. information needs, which changes based on the disease In-depth future research promises to investigate the relationship trajectory. Concurrently, their knowledge of the disease evolves between receiving social support, especially informational gradually over the course of the disease trajectory. Most patients support, and knowledge acquisition in OHCs. with OvCa have little to no knowledge of OvCa before their diagnosis due to a lack of disease awareness [26]. As the Contributions and Implications trajectory proceeds, they obtain information and gain knowledge As one of the first studies to investigate users’ knowledge through diverse sources, including OHCs [38]. However, the acquisition in the context of OHCs, this study presents several knowledge acquisition process could be extremely difficult contributions and implications to OHCs and the population of because of the lack of OvCa-related knowledge, poor quality the OvCa community. of some information available on the web, and inherent https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 12 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al characteristics of OvCa [39]. The high prevalence of questions might be biased toward the site used to collect the data. Second, associated with low- to middle-level learning objectives found the measurement of users’ learning objectives in this study was in this study further confirmed the public’s lack of awareness limited by the scope of the A&K taxonomy. Only 3 of OvCa and the community’s lack of disease knowledge. representative cognitive learning levels were selected. Such a design is based on the rationale explained in the Methods By contrast, the findings highlighted the benefits of OHC in section, but we acknowledge that users’ learning and knowledge supporting the OvCa community. Patients with OvCa and evolution was oversimplified. Knowledge acquisition is confined caregivers address their assorted information needs in OHC and to research settings. Little is known about how much the exchange information and emotional support in the community. participants learned via other information sources beyond In addition, the results based on the classification of information seeking and support within the OHC. In the future, OvCa-related topics provide insights into the information needs a complementary obtrusive approach, such as a questionnaire, of people with OvCa, such as the high demand for would help measure patients’ knowledge acquisition more treatment-related information and support. As there are multiple comprehensively. Third, this study only captures OvCa-related treatment options for OvCa, a more personalized search system topics based on the information needs of patients and caregivers. will be beneficial for providing adjusted and dynamic treatment Other types of supportive care needs, such as interpersonal or support. The findings provide implications for future health intimacy and daily living needs, were not included in the care providers, practitioners, researchers, and developers to analysis [40]. Finally, this study did not distinguish patients design personalized health information systems that will enhance with OvCa according to their disease trajectory, given the scarce knowledge acquisition and satisfy the unmet needs of people data in the NOCC. However, the literature suggests that the with OvCa. information needs of people with OvCa change with the disease trajectory [41,42]. It would be interesting to investigate whether Methodological and Theoretical Implications there is a significant effect of disease trajectory on learning In addition to the empirical and practical implications of this objectives and support in OHC. The answer to this question study, there are several theoretical and methodological may help researchers and clinicians design interventions that implications. First, this study adopted a mixed methods better support patients with OvCa along their disease trajectory. approach, which allowed us to examine both the quality and quantity of the OvCa community’s knowledge acquisition in Conclusions OHCs. Second, several coding frameworks originated from this This work is one of the first to investigate users’ participation study, such as the coding framework for OvCa-related topics in OHCs from a knowledge acquisition perspective through the and the coding framework for learning objectives. These analysis of a well-known OHC for OvCa. The results frameworks can provide future researchers with an approach to demonstrate that users use OHCs to address information needs unveil the complicated information requirements of the OvCa with different levels of learning objectives, and simultaneously, community. they can acquire various types of information and emotional support in the comments from peers. Receiving support drives Limitations and Future Directions users to pursue higher levels of learning objectives. These Regardless of its strengths, this study has several limitations. findings contribute to improving OHC designs to support the First, this study was conducted on the NOCC. Although it is a OvCa community. popular OHC for people with OvCa, the results of this study Acknowledgments This study was supported by awards from the National Library of Medicine of the National Institutes of Health (R01-LM013038). The content is the sole responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The authors would like to thank the National Ovarian Cancer Coalition for the approval of this project. Conflicts of Interest None declared. References 1. Cline RJ, Haynes KM. Consumer health information seeking on the Internet: the state of the art. Health Educ Res 2001 Dec;16(6):671-692. [doi: 10.1093/her/16.6.671] [Medline: 11780707] 2. Harkin LJ, Beaver K, Dey P, Choong K. Navigating cancer using online communities: a grounded theory of survivor and family experiences. J Cancer Surviv 2017 Dec;11(6):658-669 [FREE Full text] [doi: 10.1007/s11764-017-0616-1] [Medline: 28470506] 3. van Eenbergen MC, van de Poll-Franse LV, Heine P, Mols F. The impact of participation in online cancer communities on patient reported outcomes: systematic review. JMIR Cancer 2017 Sep 28;3(2):e15 [FREE Full text] [doi: 10.2196/cancer.7312] [Medline: 28958985] https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 13 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al 4. Yang D, Kraut R, Smith T, Mayfield E, Jurafsky D. Seekers, providers, welcomers, and storytellers: modeling social roles in online health communities. Proc SIGCHI Conf Hum Factor Comput Syst 2019 May;2019:344 [FREE Full text] [doi: 10.1145/3290605.3300574] [Medline: 31423493] 5. Wang X, High A, Wang X, Zhao K. Predicting users' continued engagement in online health communities from the quantity and quality of received support. J Assoc Inf Sci Technol 2021 Jun;72(6):710-722. [doi: 10.1002/asi.24436] 6. Wang YC, Kraut R, Levine JM. To stay or leave? The relationship of emotional and informational support to commitment in online health support groups. In: Proceedings of the ACM 2012 conference on Computer Supported Cooperative Work. 2012 Presented at: CSCW '12; February 11-15, 2012; Seattle, WA, USA p. 833-842. [doi: 10.1145/2145204.2145329] 7. Gill PS, Whisnant B. A qualitative assessment of an online support community for ovarian cancer patients. Patient Relat Outcome Meas 2012;3:51-58 [FREE Full text] [doi: 10.2147/PROM.S36034] [Medline: 23185122] 8. Anderson LW, Krathwohl DR, Airasian PW, Cruikshank KA, Mayer RE, Pintrich PR, et al. A Taxonomy for Learning, Teaching, and Assessing: A Revision of Bloom's Taxonomy of Educational Objectives, abridged edition. White Plains, NY, USA: Longman; 2001. 9. Bloom BS, Krathwohl DR. Taxonomy of Educational Objectives: The Classification of Educational Goals, Volume 1. New York, NY, USA: McKay; 1956. 10. Cartright MA, White RW, Horvitz E. Intentions and attention in exploratory health search. In: Proceedings of the 34th international ACM SIGIR conference on Research and development in Information Retrieval. 2011 Jul Presented at: SIGIR '11; July 24-28, 2011; Beijing, China p. 65-74. [doi: 10.1145/2009916.2009929] 11. Andalibi N, Haimson OL, De Choudhury M, Forte A. Social support, reciprocity, and anonymity in responses to sexual abuse disclosures on social media. ACM Trans Comput Hum Interact 2018 Oct 31;25(5):1-35. [doi: 10.1145/3234942] 12. Yang D, Yao Z, Seering J, Kraut R. The channel matters: self-disclosure, reciprocity and social support in online cancer support groups. Proc SIGCHI Conf Hum Factor Comput Syst 2019 May;2019:31 [FREE Full text] [doi: 10.1145/3290605.3300261] [Medline: 31448374] 13. Griesbaum J, Mahrholz N, von Löwe Kiedrowski K, Rittberger M. Knowledge generation in online forums: a case study in the German educational domain. Aslib J Inf Manag 2015;67(1):2-26. [doi: 10.1108/ajim-09-2014-0112] 14. Houlihan MC, Tariman JD. Comparison of outcome measures for traditional and online support groups for breast cancer patients: an integrative literature review. J Adv Pract Oncol 2017;8(4):348-359 [FREE Full text] [Medline: 30018841] 15. Wang YC, Kraut RE, Levine JM. Eliciting and receiving online support: using computer-aided content analysis to examine the dynamics of online social support. J Med Internet Res 2015 Apr 20;17(4):e99 [FREE Full text] [doi: 10.2196/jmir.3558] [Medline: 25896033] 16. Savolainen R. Requesting and providing information in blogs and internet discussion forums. J Document 2011 Sep 06;67(5):863-886. [doi: 10.1108/00220411111164718] 17. Chuang KY, Yang CC. Informational support exchanges using different computer-mediated communication formats in a social media alcoholism community. J Assn Inf Sci Tec 2014 Jan;65(1):37-52. [doi: 10.1002/asi.22960] 18. Nagler RH, Gray SW, Romantan A, Kelly BJ, DeMichele A, Armstrong K, et al. Differences in information seeking among breast, prostate, and colorectal cancer patients: results from a population-based survey. Patient Educ Couns 2010 Dec;81 Suppl:S54-S62 [FREE Full text] [doi: 10.1016/j.pec.2010.09.010] [Medline: 20934297] 19. Madathil KC, Greenstein JS, Juang KA, Neyens DM, Gramopadhye AK. An investigation of the informational needs of ovarian cancer patients and their supporters. Proc Hum Factors Ergon Soc Annu Meet 2013 Sep 30;57(1):748-752. [doi: 10.1177/1541931213571163] 20. Pozzar RA, Berry DL. Preserving oneself in the face of uncertainty: a grounded theory study of women with ovarian cancer. Oncol Nurs Forum 2019 Sep 01;46(5):595-603. [doi: 10.1188/19.ONF.595-603] [Medline: 31424458] 21. Adams NE. Bloom's taxonomy of cognitive learning objectives. J Med Libr Assoc 2015 Jul;103(3):152-153 [FREE Full text] [doi: 10.3163/1536-5050.103.3.010] [Medline: 26213509] 22. Genetic and Rare Diseases Information Center. Ovarian cancer. U.S. Department of Health and Human Services. URL: https://rarediseases.info.nih.gov/diseases/7295/ovarian-cancer [accessed 2021-11-15] 23. Cancer Stat Facts: Ovarian Cancer. National Cancer Institute. URL: https://seer.cancer.gov/statfacts/html/ovary.html [accessed 2021-11-15] 24. Hagan TL, Donovan HS. Ovarian cancer survivors' experiences of self-advocacy: a focus group study. Oncol Nurs Forum 2013 Mar;40(2):140-147 [FREE Full text] [doi: 10.1188/13.ONF.A12-A19] [Medline: 23454476] 25. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2020. CA Cancer J Clin 2020 Jan;70(1):7-30 [FREE Full text] [doi: 10.3322/caac.21590] [Medline: 31912902] 26. Reid F, Bhatla N, Oza AM, Blank SV, Cohen R, Adams T, et al. The World Ovarian Cancer Coalition Every Woman Study: identifying challenges and opportunities to improve survival and quality of life. Int J Gynecol Cancer 2021 Feb;31(2):238-244. [doi: 10.1136/ijgc-2019-000983] [Medline: 32540894] 27. National Ovarian Cancer Coalition. URL: https://nocccommunity.ovarian.org/[accessed [accessed 2021-11-15] 28. Viera AJ, Garrett JM. Understanding interobserver agreement: the kappa statistic. Fam Med 2005 May;37(5):360-363 [FREE Full text] [Medline: 15883903] https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 14 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al 29. Lombard M, Snyder-Duch J, Bracken CC. Practical resources for assessing and reporting intercoder reliability in content analysis research projects. Intercoder Reliability in Content Analysis. 2004 Oct 23. URL: https://www.researchgate.net/ publication/242785900_Practical_Resources_for_Assessing_and_Reporting_Intercoder_Reliability_in_Content_Analysis _Research_Projects [accessed 2021-11-15] 30. Devlin J, Chang MW, Lee K, Toutanova K. BERT: pre-training of deep bidirectional transformers for language understanding. In: Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies. 2019 Presented at: NAACL-HLT '19; June 2-7, 2019; Minneapolis, MN, USA p. 4171-4186 URL: http://aclanthology.lst.uni-saarland.de/N19-1423.pdf 31. HELPeR-codes / Post Support Type Prediction. GitHub. URL: https://github.com/HELPeR-codes/Post-Support-Type-Pre diction [accessed 2022-08-31] 32. Donovan HS, Hartenbach EM, Method MW. Patient-provider communication and perceived control for women experiencing multiple symptoms associated with ovarian cancer. Gynecol Oncol 2005 Nov;99(2):404-411. [doi: 10.1016/j.ygyno.2005.06.062] [Medline: 16112174] 33. Xing W, Goggins S, Introne J. Quantifying the effect of informational support on membership retention in online communities through large-scale data analytics. Comput Human Behav 2018 Sep;86:227-234. [doi: 10.1016/j.chb.2018.04.042] 34. Greene J, Hibbard JH. Why does patient activation matter? An examination of the relationships between patient activation and health-related outcomes. J Gen Intern Med 2012 May;27(5):520-526 [FREE Full text] [doi: 10.1007/s11606-011-1931-2] [Medline: 22127797] 35. Hibbard JH, Greene J. What the evidence shows about patient activation: better health outcomes and care experiences; fewer data on costs. Health Aff (Millwood) 2013 Feb;32(2):207-214. [doi: 10.1377/hlthaff.2012.1061] [Medline: 23381511] 36. Bekker HL, Winterbottom AE, Butow P, Dillard AJ, Feldman-Stewart D, Fowler FJ, et al. Do personal stories make patient decision aids more effective? A critical review of theory and evidence. BMC Med Inform Decis Mak 2013;13 Suppl 2:S9 [FREE Full text] [doi: 10.1186/1472-6947-13-S2-S9] [Medline: 24625283] 37. Zhang J. Supporting Diabetes Patient Decisional Needs Through Online Health Communities. University of California San Diego. 2019. URL: https://escholarship.org/content/qt3396035p/qt3396035p.pdf [accessed 2021-11-15] 38. Thaker K, Chi Y, Birkhoff S, He D, Donovan H, Rosenblum L, et al. Exploring resource-sharing behaviors for finding relevant health resources: analysis of an online ovarian cancer community. JMIR Cancer 2022 Apr 12;8(2):e33110 [FREE Full text] [doi: 10.2196/33110] [Medline: 35258465] 39. Chi Y, Hui V, Kunsak H, Brusilovsky P, Donovan H, He D, et al. Challenges of ovarian cancer patient and caregiver online health information seeking. Proc Assoc Inf Sci Technol 2021 Oct 13;58(1):688-690. [doi: 10.1002/pra2.530] 40. Maguire R, Kotronoulas G, Simpson M, Paterson C. A systematic review of the supportive care needs of women living with and beyond cervical cancer. Gynecol Oncol 2015 Mar;136(3):478-490. [doi: 10.1016/j.ygyno.2014.10.030] [Medline: 25462200] 41. Stewart DE, Wong F, Cheung AM, Dancey J, Meana M, Cameron JI, et al. Information needs and decisional preferences among women with ovarian cancer. Gynecol Oncol 2000 Jun;77(3):357-361. [doi: 10.1006/gyno.2000.5799] [Medline: 10831342] 42. Simacek K, Raja P, Chiauzzi E, Eek D, Halling K. What do ovarian cancer patients expect from treatment?: perspectives from an online patient community. Cancer Nurs 2017;40(5):E17-E27. [doi: 10.1097/NCC.0000000000000415] [Medline: 27454765] Abbreviations A&K taxonomy: Anderson and Krathwohl taxonomy of learning BERT: Bidirectional Encoder Representations from Transformers NOCC: National Ovarian Cancer Coalition OHC: online health community OvCa: ovarian cancer RQ: research question Edited by A Mavragani; submitted 16.05.22; peer-reviewed by K Xing, R Pozzar; comments to author 20.06.22; revised version received 08.07.22; accepted 10.07.22; published 13.09.22 Please cite as: Chi Y, Thaker K, He D, Hui V, Donovan H, Brusilovsky P, Lee YJ JMIR Cancer 2022;8(3):e39643 URL: https://cancer.jmir.org/2022/3/e39643 doi: 10.2196/39643 PMID: https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 15 (page number not for citation purposes) XSL FO RenderX JMIR CANCER Chi et al ©Yu Chi, Khushboo Thaker, Daqing He, Vivian Hui, Heidi Donovan, Peter Brusilovsky, Young Ji Lee. Originally published in JMIR Cancer (https://cancer.jmir.org), 13.09.2022. This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in JMIR Cancer, is properly cited. The complete bibliographic information, a link to the original publication on https://cancer.jmir.org/, as well as this copyright and license information must be included. https://cancer.jmir.org/2022/3/e39643 JMIR Cancer 2022 | vol. 8 | iss. 3 | e39643 | p. 16 (page number not for citation purposes) XSL FO RenderX
Journal
JMIR Cancer – JMIR Publications
Published: Sep 13, 2022
Keywords: online health community; ovarian cancer; health information needs; social support; knowledge acquisition