Introduction

Infertility is a condition characterized by “the failure to establish a clinical pregnancy after 12 months of regular, unprotected sexual intercourse”1. At present, the incidence of infertility in China is 6–15%, that is, almost 10 couples may have one pair may be infertility. As more people choose late marriage and late childbirth, this proportion may continue to expand. Despite the advancements in medical measures and assisted reproductive technology, successful pregnancy is still not guaranteed2. In addition to the objective physical reasons, psychological reasons seem to have a greater impact. Although infertility still does not get enough attention, it has become a global issue. Many regard infertility as the second top significant life stressor after the death of a loved one and divorce3.

Infertile people are highly vulnerable to multidimensional distress and desire support. Interaction with similar others and gain of practical information on infertility are two desired forms of support for infertile people4. Although supports from family members, friends and professionals are found to be welcomed, these supports have been found to be inadequate for their needs5. In addition, these supports from face-to-face communities are often unable to meet people’s psychosocial needs. Infertility problems are not obvious, person may be not able to obtain support in their face-to-face communities. These non-infertile individuals from face-to-face communities are unable to understand their situation and support from them is perceived to be less helpful and even detrimental6. They are even withdraw from social communication in face-to-face communities, as they feel ashamed7.

In this social environment, one important source for person experiencing infertility problems is online health communities (OHCs)8. OHC has emerged as a potent resource for patients seeking both anonymous and nonanonymous engagement on health problems. The role of OHCs on health has been widely revealed, including health conditions9and depression mood10,11.

In OHCs, such as Haodf.com and Guahao.com, the social support received by patients mainly consists of emotional support and informational support. These two types of support can be further categorized into professional support and peer support based on their sources.Professional support refers to assistance from healthcare professionals, including doctors, nurses, counselors, and pharmacists12.These professionals provide scientific medical advice and treatment plans for OHC users through systematic healthcare services, assisting patients in better managing their health conditions.Peer support refers to the mutual assistance among patients, primarily manifested through the sharing of treatment experiences and emotional exchanges.Generally speaking, patients can share their feelings, satisfaction, and health outcomes on the platform in OHCs. In this article, the support expressed by doctors encompasses the emotional and informational support demonstrated by physicians when patients seek assistance. In contrast, the support disclosed by peers refers to patients assessing the emotional and informational support provided by doctors through interactions and evaluations with other peers.

Social support is regarded as one of the significant benefits of OHCs13. However, on the one hand, most these studies focus on mental health diseases, few studies have looked at the social support effects of infertility. On the other hand, current studies have ignored the role of social support in user decision-making. The increasing prevalence of infertility, coupled with the lack of traditional support outlets, underscores the importance of investigating the role of online health communities as a source of emotional and informational benefits. Thus, this study attempted to solve the following research questions:

RQ1: How do informational support and emotional support expressed from doctors affect patient decision-making?

RQ2: How do informational support and emotional support disclosed from peers affect patient decision-making?

RQ3: Does social support derived from various sources exert different effects on patient decision-making?

RQ4: Whether there are individual differences (from treatment experience and emotion state) in the relationships between social support and patient decision-making?

In this study, we will first provide an overview of the research literature on online health communities, social support and online social support. Then propose research hypotheses and research model. Sections 3 and 4 present the research methodology, data analysis and empirical results. Finally, we provide both theoretical and practical implications, limitation and future directions.

Online health communities

OHCs provide a platform through which regular and ongoing communication is available to those who find face-to-face communication difficult14, and contribute to feelings of belonging, empowerment and independence through actively connecting with others15. OHCs offer a venue for people to interact with peers facing similar health problems and make it possible to exchange medical knowledge in many ways, including professional diagnosis, discussion forums and experience sharing. OHCs are a growing source of informational and emotional support for individuals dealing with health concerns as OHCs can overcome temporal and geographic boundaries16. Additionally, individuals feel more comfortable to confide personal information in a virtual environment. They can confide their feelings to and obtain support from professionals or similar others. Thus, the number of people seeking health-related support from OHCs has grown dramatically in recent years.

The emergence of OHCs provides new opportunities to study the role of online social support as these platforms have recorded social interaction among individuals. There are many studies have revealed the advantages and disadvantages of social support obtained from OHCs compared with traditional offline support17. However, most of these studies do not distinguish between general disease/knowledge and specific disease/knowledge18,19. Several studies focus on social support exchange among pregnant women, but few studies have focused on the social support among infertile patients in OHCs20,21.

Social support and online social support

Social support refers to exchange of resources between recipients and providers, reduces situation uncertainty and enhances control, and to be intended to enhance the well-being of the recipient22. Based on the resource nature, social support is a multidimensional construct and mainly covered by two types: informational support and emotional support23, which are common in OHCs. Informational support involve providing suggestions, advice, knowledge, and referrals to help others about treatment or actions24,25, while emotional support includes caring, understanding, empathy and encouragement expressions related to health concerns.

The online environment provides users with numerous opportunities for supportive activities. Users can exchange information about treatments or symptoms, express empathy or encouragement, acknowledge or validate the feelings and opinions of others, or share their own experiences26. There are many ways in which online social support can benefit users, including actively participating (i.e., providing support or receiving support) and even silently observing support exchange. Active participation in OHCs can have positive outcome, with posting messages to an online group shows greater mood improvement27,28and receiving messages of support also reports lower level of stress29,30,31. Observing the exchange of social support can influence physical health symptoms by affecting people’s sense of virtual communities, namely sense of virtual communities served as a buffer between stress and physical health symptoms.

In OHCs, such as Haodf.com, patients focus on the social support from doctors and make buying decision based on it. However, when patients need to judge a doctor’s social support behavior, they can evaluate it either from the past doctor-patient interaction texts or from the feedbacks of existing patients of the doctor. In summary, the social support received by patients can be primarily categorized into emotional support and informational support. These two types of support can be further divided into professional support and peer support based on their sources. Specifically, patients receive professional support directly from healthcare personnel, while also engaging in indirect observational exchanges of support through peer disclosures.

In our study, we include both two sources of social support and form four types: (a) informational support expressed from the doctor, (b) emotional support expressed from the doctor, (c) informational support disclosed from peers and (d) emotional support disclosed from peers.In particular, the support expressed by doctors encompasses the emotional and informational support demonstrated by physicians when patients seek assistance. In contrast, the support disclosed by peers refers to patients assessing the emotional and informational support provided by doctors through interactions and evaluations with other peers.

The influence of social support and patient decision-making

The most important reason for people suffering from diseases to engage in OHCs is to access medical information and obtain social support32. Informational and emotional support are the key dimensions of social support in OHCs. Informational support and emotional support can satisfy individuals’ unmet needs offline and prompt users to significantly engage in OHCs33. Based on literature of online engagement, meeting informational and emotional needs will evoke users’ desire to involve in OHCs34. Members can be engaged in OHCs if informational and emotional benefits are to be realized35. In the doctor-patient communities (such as Haodf.com and Guahao.com), doctors provide social support by providing services with a fee and patients provide social support by sharing treatment experience publicly. A user will engage in OHCs by buying service of doctors if he feel his informational and emotional needs can be met. As a result, we hypothesize that the informational and emotional support expected to obtain from a doctor in OHCs can create a strong patient decision-making.

Hypothesis

a: Informational support expressed from a doctor positively influences patient decision-making in OHCs.

Hypothesis

b: Emotional support expressed from a doctor positively influences patient decision-making in OHCs.

Hypothesis

a: Informational support disclosed from peers positively influences patient decision-making in OHCs.

Hypothesis

b: Emotional support disclosed from peers positively influences patient decision-making in OHCs.

The role of patient characteristics

Evidence indicates that the emphasis on informational versus emotional support may vary depending on the type of health issue focus upon. For instance, users in OHC that focus on breast cancer are more likely to seek and provide emotional support, while users of a prostate cancer OHC are more likely to seek and provide informational support36In addition, gender has also influence, with women are more likely to engage in emotional support than men37,38. In this study, we focus on treatment experience and emotion state of patients.

Experience is a critical factor that influence key constructs and relationships39,40, as experience increases individual’s understanding about environment and situation. Based on the self-perception theory, an individual will adjust his perceptions by observing the behavior outcome and the adjusted perceptions will affect subsequent behaviors41. High-experience users will be able to assess and evaluate at a deeper level and they tend to refocus their attention from social concerns to service quality and economics benefits42. It is expected that user experience will affect their perception on the relative importance of the social exchange. People with more treatment experience tend to perceive more control and experience less emotional strain. Therefore, we expect that the treatment experience can moderate the relationship between social support and decision-making.

Hypothesis

a: The relationship between informational support and patient decision-making will be stronger for users with high levels of treatment experience than that for users with low levels of treatment experience.

Hypothesis

b: The relationship between emotional support and patient decision-making will be stronger for users with low levels of treatment experience than that for users with high levels of treatment experience.

Emotion state also act as a moderator. Individuals experiencing high pleasure have a better ability to deal with information and complexity successfully, and thus are more aware of stimuli43, which may facilitate confidence in decision-making44. Positive emotion state strengthens the effects of service quality and satisfaction45. In addition, emotion can affect the cognitive of individual, therefore directly influence their information processing46and decisions47. Thus, patients in different emotion states may place different importance on social support.

Hypothesis

a: Emotion state positively moderates the relationship between informational support and patient decision-making. That is, information support has a greater influence on decision-making for patients with better emotion state.

Hypothesis

b: Emotion state negatively moderates the relationship between emotional support and patient decision-making. That is, emotional support has a smaller influence on decision-making for patients with better emotion state.

Methods

Data sources

We select Haodf.com as our data source. Haodf.com was founded in 2006 and is one of the leading OHCs in China. It has many high-quality doctors. As of July 2023, Haodf.com has collected the information of 910 thousands of doctors from more than 10 thousands of regular hospitals in China. Among them, 270 thousands of doctors provide online medical services directly to patients. On Haodf.com, patients can inquire the hospital and doctor’s professional expertise, patient rating and treatment experience on the doctor’s individual homepage. Based on that information, patients can choose a more suitable doctor according to their own condition. Moreover, patients can also buy doctors’ services and communicate with doctors online. Doctors will give reasonable suggestions or treatment plans based on the patient’s condition information and consultation information without time cost and traffic cost.

As text-picture consultation is the main type of services on Haodf.com and interactions between doctors and patients is recorded and public, it is helpful for us to label informational and emotional support. A python-based web crawler was developed to collect data from Haodf.com. Based on our research design, we collected information of all doctors who can treat infertility and interactions between these doctors and latest 20 patients on September 2022. We also collected the latest 20 patient treatment experience of these doctors. Finally, 2,989 doctors with 39,200 doctor-patient interactions and 15,960 treatment experiences are included. On December 2022, we collected information for all new patients of 2,989 doctors and finally 25,100 patient were included.

Measures

The dependent variable is the increment in number of patients for doctor i between September and December 2022 (IoPi). For the independent variables, two sources of social support are included in our research: social support expressed from the doctor and social support disclosed from peers. We measure social support expressed from the doctor based on the interactions of the doctor and his latest 20 patients, and measure social support disclosed from peers based on his latest 20 patient treatment experience of each doctor. When a patient makes decision, he need to perceive informational support and emotional support based on the existing public data that comes from the doctor-patient interactions and patients’ treatment experience on Haodf.com. We can hypothesize that when a patient come to a doctor’s homepage, he can browse 20 doctor-patient interactions and 20 treatment experiences at most and make a decision. Therefore, informational support and emotional support are measured based on the latest 20 interactions and the latest 20 treatment experiences on September 2022 through text mining.

Informational support and emotional support are extracted from the doctor-patient interaction’s text information and patient treatment experience through text mining. For informational support, the total number of professional words used by the professional is measured. These professional words include suggestions and guidance on disease, drug, exercises, etc. The emotional support is measured using the numbers of emotional words used by the doctor. For example, come on, don’t give up, you will be fine, etc. For the treatment experience, the rules are similar, except that these professional and emotional words come from patients’ reports. In order to ensure the quality of labeling, two professional assistants help label 1000 text samples of this study independently. Through the cross check test, they analyze the reasons for inconsistent parts and reach a consensus. Following previous studies, we use Jieba word segmentation tool and TF-IDF algorithm to segment word and find keywords48,49. Then random forest algorithm is used to classify 750 training texts and 250 test texts. For both informational support and emotional support, the performance of random forest algorithm meets the classification requirement (Accuracy > 90%, Precision > 90%). Thus, the classifier is used to extract both informational support and emotional support from the unlabeled doctor-patient interaction text. Based on the above method, we get informational support and emotional support expressed from doctors (ISEPi and ESEPi) and informational support and emotional support disclosed from peers (ISDPi and ESDPi).

For the moderate variables, treatment experience can be obtained based on patient complain on Haodf.com directly. We measure all the treatment experience of these new patients between September and December 2022 and then take the average value (TreatExpi). The sentiment analysis application program interface from Baidu AI-NLP is used to measure emotion state of patients and then the average value is included (EmoStai), which is widely used in previous studies9,21.

In order to eliminate the influence of other factors, control variables including the number of patients online (#Patienti), the recommendation (Recommendationi), hospital level (Hleveli), doctor medical title (Mtitlei) and service price (Serpi). The detailed definitions for the main variables included in this study are shown in Table 1.

Table 1 Variable definitions.
Full size table

Analysis

We use regression model to analyze data. First, we verify the effects the informational support perceived and emotional support perceived on patients’ choices (Eq. 1). Second, we add the moderate variables to reveal the moderating effects of treatment experience and emotional state (Eq. 2). The equations are presented as follows.

$$Io{P_i}={\beta _0}+{\beta _1}ISE{P_i}+{\beta _2}ESE{P_i}+{\beta _3}ISD{P_i}+{\beta _4}ESD{P_i}+{\beta _{5 - 9}}Control{s_i}+\varepsilon$$
(1)
$$\begin{gathered} Io{P_i}={\beta _0}+{\beta _1}ISE{P_i}+{\beta _2}ESE{P_i}+{\beta _3}ISD{P_i}+{\beta _4}ESD{P_i}+{\beta _5}TreatEx{p_i}+{\beta _6}EmoSt{a_i} \hfill \\ {\text{ }}+{\beta _7}ISE{P_i} \times TreatEx{p_i}+{\beta _8}ESE{P_i} \times TreatEx{p_i}+{\beta _9}ISD{P_i} \times TreatEx{p_i}+{\beta _{10}}ESD{P_i} \times TreatEx{p_i} \hfill \\ {\text{ }}+{\beta _{11}}ISE{P_i} \times EmoSt{a_i}+{\beta _{12}}ESE{P_i} \times EmoSt{a_i}+{\beta _{13}}ISD{P_i} \times EmoSt{a_i}+{\beta _{14}}ESD{P_i} \times EmoSt{a_i} \hfill \\ {\text{ }}+{\beta _{15 - 19}}Control{s_i}+\varepsilon \hfill \\ \end{gathered}$$
(2)

Results

Description results

Descriptive statistics of the main variables are shown in Table 2. Python is used to analyze the data.

Table 2 Description of variables.
Full size table

Empirical results

The empirical results in Table 3 suggest that both informational support and emotional support expressed from the doctor positively affect patient decision-making (β = 0.050, p < 0.001; β = 0.017, p < 0.001, respectively). The results support H1a and H1b: Informational support/Emotional support expressed from a doctor positively influences patient decision-making in OHCs. In addition, both informational support and emotional support disclosed from the peers also positively influence patient decision-making (β = 0.025, p < 0.001; β = 0.014, p < 0.001, respectively). Thus, H2a and H2b are supported: Informational support/Emotional support disclosed from peers positively influences patient decision-making in OHCs.

Table 3 Empirical results.
Full size table

We find that the coefficient of treatment experience and emotion state are also significant, which means that the moderating effects can be checked. The results in Model 4 in Table 3 supports H3a: The relationship between informational support and patient decision-making will be stronger for users with high levels of treatment experience than that for users with low levels of treatment experience (β = 0.019, p < 0.01; β = 0.006, p < 0.05, respectively). H3b is also supported: The relationship between emotional support and patient decision-making will be stronger for users with low levels of treatment experience than that for users with high levels of treatment experience (β=−0.013, p < 0.05; β=−0.005, p < 0.05, respectively). For the moderating effects of emotion state, the results indicate that patients with different emotion state might not have different value on informational support. Thus, H4a is not supported. However, we find that emotion state significantly and negatively moderates the relationship between emotional support and patient decision-making (β=−0.010, p < 0.01; β=−0.007, p < 0.05, respectively). Namely, emotional support has a smaller influence on decision-making for patients with better emotion state. Therefore, H4b is supported.

Robustness checks

To alleviate the potential model setting bias, we performed an additional robustness check to further ensure the reliability of our findings. We include theses doctors who have more than 50 patients into the samples. As doctors experience more patients, their service quality tend to be stable. Thus, the measures of social support are also more accurate. By using the new samples, we reanalyze the empirical model and the results are consistent with the main results. Therefore, we can conclude that our findings are robust. The robustness checks results are shown in Table 4.

Table 4 Robustness checks.
Full size table

Discussion

Key findings

Given the importance of online social support for infertile patients, many OHCs seek to promote social support exchange effectively and Social support is regarded as one of the significant benefits of OHCs9,13. However, current studies mainly focus on mental health diseases, and few studies have looked at the social support effects of infertility. Moreover, the role of social support in user decision-making is ignored. Our study offers valuable insights into underscoring the importance of investigating the role of online health communities as a source of emotional and informational benefits.

Overall, the empirical results support most of the hypotheses. Two sources of social support (expressed from doctor versus disclosed from peers) and two kinds of social support (informational support versus emotional support) have played their roles in the patient’s decision-making. Regarding the role of patient characteristics, both treatment experience and emotion state demonstrate moderating effects on the relationship between social support and decision-making.

The heterogeneous role of social support types and sources

The role of both informational and emotional support are examined in this study. Although previous literature has demonstrated the effectiveness of social support in online environment50, our finding reveals the its effective role in the patient decision-making. Moreover, we include two types and two sources of social support and find their heterogeneous role.

For the social support types, although both informational and emotional support have been found positively related with patient decision-making, the effect of informational support is larger than that of emotional support. This indicates that informational support plays a more important role in patient decision-making process. For the social support sources, we include social support expressed from doctor directly and social support disclosed from peers indirectly. Although results show that different sources of social support have significant positive effects on decision making, the coefficients of social support expressed from doctor is larger. Therefore, when faced with a choice, the patient will value informational support and social support expressed from doctors more.

The moderating role of patient characteristics in decision-making

Based on the elaboration likelihood mode, person characteristics will influence their attitude-formation process51. Evidence also indicates that the emphasis on informational versus emotional support may vary depending on the type of health issue focus upon36and gender37,38. Therefore, we include treatment experience and emotion state as moderating variables based on our research design.

As experience increases individual’s understanding about environment and situation, it is a critical factor that influence key constructs and relationships. Our results show that for patients with higher treatment experience, information support will play a greater role of in decision. In addition, emotion can affect the cognitive of individual, therefore directly influence their information processing and decisions. The results show that for patients with better emotion state, the emotional support will have a smaller influence on decision-making.

Theoretical implications

Our study provides the following theoretical contributions. First, this study is one of the first to reveal the role of social support on decision-making behavior in the online health context, where effective social support and knowledge sharing are important to ensure medical service quality and safety52,53. Most previous studies mainly focused on the effectiveness of social support on health state9,32, such as mental health, very few studies have examined the role of online social support in patients’ purchase decision-making. This study extends the social support theory to decision behavior, and we verified that both informational support and emotional support play roles in decision-making.

Second, this study contributes to existing studies by focusing on the social support among infertile patients in OHCs. Social support is important for improving patient health54,55,56, and this study contributes to a deeper understanding of the role social support plays in individual health. Although there are many studies have revealed the advantages and disadvantages of social support obtained from OHCs compared with traditional offline support17. However, most of these studies do not distinguish between general disease/knowledge and specific disease/knowledge18,19,20,21. Only several studies focus on social support exchange among pregnant women9,50. Our results show that social support play a particularly important role in decision-making for infertile patients.

Third, we contribute to the literature on social support by examining the role of different sources of social support. Prior studies primarily focus on the peer side or professional side to investigate the social support effect from one side57. Our study provides evidence that social support from both peers and professionals can benefit decision-making process. In particular, our findings further reveal that the social support from professional play a more important role compared with social support from peers. In addition, we also find that patients value informational support more than emotional support.

Fourth, this study extends our understanding of online social support in OHCs by conceptualizing the moderating effects of patient characteristics (treatment experience and emotion state) on the relationships between social support and decision-making. Based on the elaboration likelihood mode, person characteristics will influence their attitude-formation process51. Evidence also indicates that the emphasis on informational versus emotional support may vary depending on the type of health issue focus upon and gender36,37,38. Therefore, we include treatment experience and emotion state as moderating variables based on our research design. Our results show that for patients with higher treatment experience, information support will play a greater role of in decision. We also find for patients with better emotion state, the emotional support will have a smaller influence on decision-making.

Practical implications

From a management perspective, our research has important implications for service delivery practices and patient health improvement in an online environment. For physicians, understanding the relationships between social support and decision-making process helps doctor provide service better. Our findings emphasize the importance of providing both informational support and emotional support and suggest that doctors need to emphasize the provision of emotional support, especially for infertile patients. In particular, we find a potential role of social support disclosed form peers by posting treatment experience, which suggests that doctors need maintain patient feedbacks. In addition, as patient characteristics play moderating role, doctors need to pay attention to patients’ emotion state and experience when provide medical services.

For patients, social support not only significantly improves their condition and quality of life but also enhances their ability to cope with stress and challenges, thereby promoting better mental health. The online environment has expanded the avenues through which patients can access support, moving beyond traditional face-to-face interactions. It has created a space where patients can express themselves, share experiences, and seek assistance.Therefore, patients can proactively seek social support from various sources, such as online health communities (OHCs) and family members, based on their individual needs, thereby enhancing their physical and mental well-being.

For OHCs managers, they should align management and incentive policies with the type of diseases. Patients with different disease have a different way to satisfy regarding social support. In addition, OHC platform managers should implement measures to encourage doctors provide more informational support and emotional support. For instance, set up a reminder mechanism. At the same time, it is essential to continuously optimize platform features to better facilitate information exchange and emotional interaction between healthcare providers and patients. In summary, this study provides valuable insights and practical guidance for physicians, online health community managers, and patients, helping them to better utilize online health community resources to enhance the quality of medical services and improve patient health outcomes.

Limitations and future directions

Our work can be extended in several directions. First, patient decision making is a dynamic process, and this study used static data, cannot capture the dynamic effects of social support on patient decision-making over time and future researchers could design longitudinal studies to replicate the research findings. Second, our data is collected from Haodf.com, vwhose user groups have certain limitations. whose user groups have certain limitations. Other online platforms can be used to conduct comparative studies to increase the generalization and diversity of the research in the future. Third, this study primarily relies on available text data when analyzing patient decision-making within OHCs. However, due to the anonymity and privacy protection measures in OHCs, we are unable to obtain patients’ demographic characteristics. These factors are crucial for understanding patient behavior, the decision-making process, and the impact of social support. Our study may not have fully considered these factors in model construction and analysis, potentially limiting the depth and breadth of the findings. Future research could collaborate with other research institutions or healthcare organizations to share anonymized non-sensitive data, thereby expanding the data foundation of the study and enabling a more comprehensive analysis of patient behavior and decision-making processes.

Conclusions

This study focuses on the impact of online social support on the decision-making of infertility patients. Through text mining, it examines how different types and sources of social support activities contribute to understanding the decision-making behavior of patients in OHCs. Our results show that the effect of informational support is larger than that of emotional support. Meanwhile, when faced with a choice, the patient will value informational support and social support expressed from doctors more. For patients with higher treatment experience, information support will play a greater role of in decision. For patients with better emotion state, the emotional support will have a smaller influence on decision-making.This study contributes social support theory and OHC studies, and helps the management and design OHCs.