The present study shows that SML can provide novel insights into the experiences and perceptions of SCD stakeholders, including patients, family and friends, caregivers, and HCPs. To the best of our knowledge, this is the largest SML study ever conducted on SCD, in terms of both number of analyzed posts and included countries, as only one previous SML on SCD in the US that analyzed 50 posts has been published to date.25 SML findings complement information provided by other real-world data sources and make it possible to detect issues that are more difficult to record on studies using traditional research approaches. For instance, while data on patient’s feelings or important experiences can be spontaneously shared in real time on the chosen platform, traditional research methodologies require the presence of the HCP or the use of pre-specified platforms, hampering the collection of data on relevant aspects to patients. Indeed, the use of social media as a source for collecting information on symptoms and disease impact from the patient’s perspective was proposed by the Food and Drug Administration (FDA) in 2018.26 Moreover, the FDA is exploring the value of social media mining for earlier detection of rare and serious adverse events.27
Since SML studies analyze information from public sources on social media platforms, they avoid several logistical challenges of traditional research, allowing for accelerated and cost-effective data collection.28 However, they pose unique ethical challenges, since participants do not formally consent to their data being used in the study.29 Despite the absence of current guidelines on consent or anonymity for social media research, it is recommended that data collected be used to answer specific research questions and presented in an anonymous manner.30 As SML studies become more common29,31−33, more guidelines to conduct this type of studies are expected to be available.
Twitter was, by far, the most popular social media platform for information sharing between SCD stakeholders. In recent years, Twitter has become one of the most important social media platforms in healthcare communication, with an increasing number of patients and HCP professionals sharing a wide range of experiences there.34,35 Most SCD patients were young and female. The age range of the patients did not come as a surprise since young people are the main users of social media platforms.36 However, the gender share observed here does not entirely correspond with the overall gender distribution in social media platforms, where users are predominately male,37 or with the prevalence of the disease itself, where the gender distribution is equal.38 The gender disparity in conversations on SCD might be, at least in part, due to gender stereotypes, such as men having to be strong and not showing vulnerability. In fact, male patients mentioned in some conversations how hard it was to share their struggles with SCD due to these stereotypes. This finding is in line with previous studies showing that women express more personal issues in social networks than men.39
The study highlighted the desire of patients and caregivers to discuss SCD-related topics and increase public awareness and access to information. Peak of posts was in June 2020, coinciding with the World Sickle Cell Awareness Day (19th June). At least half of the conversations were around the patient journey, with symptoms being the most discussed topic. Within the context of symptoms, conversation manly focused on pain (acute episodes and pain in general) and acute crises. Pain was not only the most discussed symptom but also the most mentioned factor having an impact on the physical domain of their QoL. In line with this, the Sickle Cell World Assessment Survey (SWAY) of 2145 SCD patients showed that, in high income countries, SCD patients had pain 2.8 days per week on average and a median of 4 VOCs during the previous year.9 Almost a quarter of these VOCs were managed at home due to a previous poor experience at the hospital,9 highlighting the need to improve management of VOCs.
The most frequently mentioned unmet needs cited in conversations were lack of SCD awareness by HCPs and society and lack of empathy from HCPs, which has been reported in previous studies.40–42. In minority ethnic groups, poorer pain management, less respectful behavior, and undertreatment have been identified,43,44 probably reflecting the underlying structural racism that persists in most western societies.16 The stigma of SCD patients can have a negative impact on the psychological, physical, and social wellbeing and impairs healthcare interactions and clinical outcomes.41,45 In April 2019, a 21-year-old black patient in the early stages of a sickle cell crisis died at a hospital in the UK, after he was denied oxygen by a HCP. We found mentions by stakeholders in social media platforms criticizing this incident. Patients also mentioned to be perceived as drug addicts and having their suffering questioned. Acknowledging the stigma and racial bias in SCD is essential to implement education and training programs among HCPs. In this regard, actions to reduce the impact of racism on patients with SCD have been recently proposed in the US and UK.16,46
Fatigue was also a frequently mentioned symptom in social media conversations, confirming the impact of this symptoms on patients QoL previously suggested.9 Studies have further showed that fatigue is highly correlated with depression, anxiety and stress,47 and interferes with daily activities,48 which may justify its evaluation as an efficacy endpoint in clinical trials.9
Treatment was the second most discussed topic of the patient journey in the conversations. Blood transfusions was most frequently mentioned treatment, which was often associated with efficacy, a feature also attributed to stem cell transplants. Stem cell transplant was the only treatment positively perceived in more than half of the conversations. Stem cell transplantation is so far the only curative option in SCD.14 Stem cell transplantation protocols are improving rapidly, due to advances in gene editing techniques that help in the genome modification of hematopoietic stem cells.49 One of these techniques is CRISPR, which enables the precise replacement of a specific region of DNA. Some CRISPR-based treatments for SCD have been already approved by the regulatory agencies and others are still under research.50 In conversations in Austria and the UK, patients shared that they had been cured with gene therapy, mentioning CRISPR. Gene therapy requires specialised centres for patient care, and therefore accessibility of gene therapy across the globe, especially in low-income areas with a high prevalence of the disease, remains a challenge that needs to be addressed.51,52
Discussion on QoL took up almost half of the conversations, with emotional impact as the most affected dimension. Main reasons for the impairment of psychological wellbeing were repeated hospital visits, lack of awareness and support from HCPs and the public and having to deal with symptoms like pain. Results from the SWAY study showed that 60% of patients stated that SCD caused a high negative impact on emotions; depression and anxiety were reported by 39% and 38% of patients, respectively.9 The SWAY study also showed that only one third of patients were receiving professional emotional support, while 62% had a desire to receive any or more of this type of support.9 In our study, among conversations on the emotional impact of the disease, mentions to negative feelings were frequently detected, but also, albeit less frequently, depression, anxiety and suicidal thoughts were discussed. The prevalence of depression in SCD has been found to be up to 5 times higher than in the general population, with higher levels of depression associated with lower QoL.53 Studies using validated patient-reported outcomes (PROs) have showed that QoL was impaired in patients with SCD.9,54−56 Importantly, patients who claimed to have suffered prejudice in the past had significantly worse QoL than those who did not.57 Other factors associated with worse QoL in SCD were older age, female gender, rural residence, low family income, disease-related complications, hospital admissions, and severity and frequency of VOCs.56,58,59
In the present SML study, financial issues were also mentioned as affecting their QoL and being an unmet need, although to a lesser extent than emotional, physical, and social aspects. Most of the countries where conversations took place have healthcare systems with a wide coverage of the disease, resulting in a lower cost for patients compared to other regions in the world60. However, some mentions of patients having to pay for life-saving medications themselves were found in France and the UK. Furthermore, SCD has indirect costs to patients, owing to high unemployment rates and difficulties keeping a job due to absenteeism and productivity loss,9,61 frequently caused by VOCs.56 A considerable number of patients mentioned that SCD affected their work or school and a few stated that they even suffer discrimination at the workplace.
Regarding complications due to SCD, a higher risk of COVID-19 was a key concern in conversations. A 4-fold increased risk for COVID-19–related hospitalization and a 2.6-fold increased risk for COVID-19–related death were observed in a large cohort of SCD patients in the UK using QResearch (a database covering approximately 18% of the UK population).62 Another study conducted using the same database also showed that SCD was a risk factor for severe COVID-19 outcomes even after one or two doses of COVID-19 vaccination.63 Future studies will show the risk for severe COVID-19 with the new strains in this population.
Overall, our findings provide a comprehensive overview of the issues of interest and concern to SCD stakeholders, mainly patients. SML presents as a relevant tool to explore on first-hand the feelings, perceptions, and experiences of people affected with SCD when they spontaneously choose to share these issues. Studies using this methodology are expected to increase, since they provide complementary information to that obtained using traditional research methods on the patient journey, patient QoL and unmet needs, among other aspects.
Limitations
The present study has several limitations inherent to social media research. First, self-selection bias might have occurred since participants in the social media platforms could have specific demographic or clinical characteristics and the willingness to participate in these platforms. Even though this might have hindered the generalizability of the results, the current study included a wide range of social media platforms and a high number of European countries, which might have mitigated this potential bias. Second, it could also be argued that the veracity of the information shared on posts was not verified; however, the benefit of sharing false information in this context seems unlikely, since the aim of posting health-related information is usually to share feelings and experiences within the community, increase awareness and find peer support. Lastly, the unstructured nature of the data also led to variation in the sample size for each study variable; this, together with the variety in number of posts from different countries, did not allow direct comparisons of conversation content between countries.