High prevalence of persistent COVID‐19‐related health anxiety and social restriction in patients with haematological disorders

During the early stages of the COVID-19 pandemic, there was a high prevalence of health-related anxiety. Patients with haematological disorders were realised to be a highly vulnerable group1, 2 and were encouraged to shield and restrict social contact. However, widespread vaccination and the reduced pathogenicity of the more recent Omicron strain have greatly reduced the mortality3, 4 and morbidity,5 even in haematology patients. As such, in the United Kingdom, national guidance on lockdowns and social distancing has returned to the pre-COVID-19 paradigm. Despite this, it is our experience that many haematological patients continue to exercise extreme caution and restriction in their social contacts, and the prevalence of ongoing COVID-19-related health anxiety and social avoidance behaviours in haematology patients remains unclear. Here, we conducted a cross-sectional survey to understand the persistence of COVID-19-related health anxiety and lifestyle restriction in patients with haematological conditions, 3 years from the start of the COVID-19 pandemic. All 604 patients with haematological disorders under our care at a single centre were invited to complete an online or paper questionnaire between 14th March and 29th April 2023 (Supplementary Methods). Patients were prompted again 4 weeks after the initial invitation to submit their responses before the closing date. Patients self-reported their demographics, physical and mental health diagnoses, and COVID-19 history and completed Generalised Anxiety Disorder-7 (GAD-7) screening. In addition, patients rated the impact of COVID-19 on their mental health (MH; scale of 0–10) and were asked to what extent (on a 5-point scale, from strongly disagree to strongly agree) concern due to COVID-19 impacted health anxiety or social avoidance in seven domains of everyday living (social interactions, visits to pubs/restaurants, shopping, public transport, travel abroad, work, and physical activity). For each domain, patients were deemed to have persistent COVID-19-related health anxiety or social avoidance if they somewhat or strongly agreed that concern due to COVID-19 impacted these factors. In total, 233 patients with haematological disorders completed the survey (Table 1). Complete responses were obtained for all of the mandatory questions. Responses to the two optional free-text questions were not analysed. The median age was 71 years and 41% were female (compared with a median age of 67 years, and 43% female amongst all invited patients). The cohort comprised patients with lymphoid (73%), plasma cell (16%), myeloid (7%), red cell (1%) and other (3%) haematological conditions. Compared with all invited patients, myeloid patients were relatively under-represented in those responding, having made up 13% of those contacted, with no other marked differences in response rates in the other diagnostic groups. The mean EQ-5D index value of the cohort was 0.742 (similar to a mean index of 0.771 in an age and sex-matched cohort from the general population). 61% reported additional comorbidities. 129 (52%) patients reported a previous history of COVID-19 infection, of which 9 (7%) were hospitalised and 26 (20%) had persistent COVID-19-related symptoms. 10% reported having a previous MH disorder prior to the COVID-19 pandemic, including depression (3%), anxiety (3%) or both anxiety and depression (<2%). We studied patient-reported social avoidance and its impact on MH due to concern about COVID-19 in this cohort of patients. In total, 50% of patients reported at least 1 avoidant behaviour across the 7 domains, whilst 19% reported this in 5–7 domains (Figure 1A). The most common persistent changes in behaviour due to concern about COVID-19 were reduced social interactions (42%), use of public transport (34%) and visits to pubs/restaurants (30%; Figure 1B). When asked to rate the impact of COVID-19 on MH, 51% reported 0/10 (no effect), whilst 12% reported 7/10 or greater. MH impact scores positively correlated with the number of social avoidance behaviours (r = 0.35, p < 0.0001). On a standardised scale from 0 (death) to 1 (perfect health), patients with at least one avoidant behaviour had a lower EQ-5D index value compared to those without (0.703 vs. 0.781; p = 0.0063); this difference would be considered to exceed the minimal clinically important difference (MCID) threshold for EQ-5D.6 We assessed associations with demographic and physical health factors. There was no difference in MH impact or avoidance behaviours due to COVID-19 by sex, number of cohabitants or employment, and no correlation with age (p > 0.05); however, there was a greater impact on MH in non-White compared to White ethnic groups (median 4/10 vs. 0/10; p = 0.0323). COVID-19-related MH impact did not correlate with number of comorbidities and did not differ by chemotherapy status (p > 0.05); however, patients having active chemotherapy had greater social avoidance compared with those with previous historical chemotherapy (median of 2 vs. 0 domains, respectively; p = 0.0192). Next, we examined the association between MH and COVID-19-related health anxiety. In total, 47% of patients reported ongoing health concern due to COVID-19 (Figure 1C); within this group, there were a greater proportion of patients also reporting some impact on MH (67% vs. 35%; p < 0.0001) and greater social avoidance (median 3 vs. 0 domains; p < 0.0001). The mean GAD-7 score was 2.7 (similar to a mean GAD-7 score of 2.9 in a sex-matched cohort from the general population7). Whilst the mean GAD-7 score was greater in respondents with an avoidance behaviour compared with none (3.8 vs. 1.7; p < 0.0001), this did not exceed reported MCID thresholds for GAD-7.8, 9 However, EQ-5D anxiety/depression mean raw scores in patients with ≥1 and 0 avoidance behaviours approximated to “slight” and “no” anxiety or depression respectively (1.71 vs. 1.35; p = 0.0004), suggesting an association between increased anxiety and social avoidance that may exceed MCID. Consistent with this, those with a positive GAD-7 screen for at least mild anxiety (score ≥ 5) reported greater concern over health due to COVID-19 (69% vs. 45%; p = 0.0027) and were 3.2 times more likely to have at least one avoidant behaviour (95% CI: 1.6–6.2, p = 0.0007), such as reduced social interactions (59% vs. 42%; p = 0.0306) and use of public transport (54% vs. 36%; p = 0.0251). The impact of COVID-19 on MH was greater in patients with a positive versus negative GAD-7 screen (median 4/10 vs. 0/10, respectively; p < 0.0001; Figure 1D, left panel), and numerical GAD-7 score had a positive correlation with MH impact (r = 0.44, p < 0.0001). Patients with a history of an MH diagnosis prior to the pandemic suffered a greater impact of COVID-19 on MH (median 5/10 vs. 0/10; p = 0.0015). Taken together, these results suggest that previous MH diagnoses or current generalised anxiety are associated with ongoing COVID-19-related MH impact and social avoidance in haematology patients. We evaluated associations with previous COVID-19 history. Patients with a social avoidance behaviour were 41% less likely to have a previous history of COVID-19 infection than those without ongoing social restriction (95% CI: 0.35–0.98, p = 0.0430). Of patients reporting previous COVID-19 infection, those with persistent post-COVID-19 symptoms recorded a higher impact of COVID-19 on their MH than individuals now asymptomatic (median 3/10 vs. 0/10 impact; p < 0.0001; Figure 1D, right panel) and were 2.6 times more likely to have an avoidant behaviour (95% CI: 1.1–6.3, p = 0.0376). Ongoing post-COVID-19 symptoms were associated with greater health concern due to COVID-19 (76% vs. 35%; p = 0.0003) and avoidance of public transport (54% vs. 27%; p = 0.0133), pubs/restaurants (44% vs. 17%; p = 0.0052) and physical activity (42% vs. 8%; p = 0.0006). These results suggest that whilst previous COVID-19 infection is generally associated with reduced persistence of social restriction, those with ongoing post-infection symptoms are more likely to maintain greater social precautions. Our study highlights ongoing COVID-19-related anxiety and behavioural change in patients within our haematology service. There was a high prevalence of anxiety and depression in haematology patients at the start of the COVID-19 pandemic.10-12 In 2020, the introduction of precautionary social restriction was paramount to reducing risk, and measures were shown to reduce anxiety by decreasing fear of contracting.13 We5 and others4 have shown that the absolute risks have since declined considerably in haematology patients. Yet, in our cohort, half of the patients had at least one ongoing behavioural change. Consistent with associations of anxiety due to COVID-19 in the general population, we observed greater social restriction in patients with a positive GAD-7 screen14 or persistent post-COVID-19 symptoms.15 There are some limitations of our analysis. Firstly, whilst we captured a range of haematological diagnoses, patients were recruited from a single centre and those who chose to participate were self-selecting; therefore, our cohort may not be representative of the wider group. Secondly, data were self-reported and so the accuracy of clinical correlates could not be verified. Thirdly, whilst we highlight associations with ongoing social restriction, our cross-sectional survey data cannot establish causal relationships with these variables. Nevertheless, our approach enabled us to capture the subjective experiences of patients and correlate them with a wide range of clinical variables. In conclusion, our data reveal a high prevalence of ongoing COVID-19-related health concerns and social avoidance in haematology patients. Given that the absolute risks to patients from contracting COVID-19 have fallen considerably since the start of the pandemic, these findings suggest the importance of good communication by clinicians in order to empower patients to make balanced decisions about the potential relaxation of their social restrictions. J. Willan initiated the project. The survey was designed by J. Willan, G. Agarwal, S. Varghese and M. Francis, and distributed by J. Willan. J. Willan and G. Agarwal analysed the data and wrote the manuscript. All authors read and approved the final manuscript. The EuroQol-5D questionnaire was used with permission from the EuroQoL Research Foundation. Thanks to Catriona Gilmour Hamilton for help with the questionnaire design. Many thanks to all the patients for their time taken to participate in this study. The authors received no specific grant from any funding agency in the public, commercial or not-for-profit sectors. The authors declare that we have no relevant conflicts of interest. All required local approvals for performing this work were obtained. Patients responding to the survey remained anonymous, were explained the rationale of the work and consented that responses would be disseminated through publication at the time of data collection. Data S1: Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.