Introduction

Multiple myeloma (MM), a haematological cancer with a high incidence [1], is a recurrent and malignant clonal disease of bone marrow plasma cells [2, 3]. An epidemiological survey showed that the worldwide incidence of MM is estimated to be 6–7/10,000 [4]. Most of the MM patients were old people, and over 50% of patients are over 65 years old [5]. The development of the ageing population in many developed countries and the diagnostic capacity have increased global incidence [6]. Thus, researchers have focused more on the symptomatic management of MM due to this increasing incidence.

The MM is a severe disease. A study by Sidhu and Homsi reported that MM ultimately causes organ damage, including anaemia, renal problems, hypercalcaemia, and osteolytic bone lesions [7]. Especially, according to Mathew et al., approximately 70% of MM patients suffer from pain because of osteolytic bone lesions [8], with pain being one of most common symptoms [9]. In addition, the MM is prone to infection and thrombosis [3]. Therefore, the MM can seriously affect the quality of life and survival rate of patients [10]. Recently, to delay the progression of the disease and to prolong the survival of MM patients, chemotherapy has been used to treat newly diagnosed patients and transplant-ineligible patients in clinical settings [11].

Compared to patients not receiving chemotherapy, the symptoms of patients undergoing chemotherapy are more complex. MM patients undergoing chemotherapy not only suffer from the symptoms of the disease but also experience various adverse effects with chemotherapy [12]. For instance, Nathwani et al. reported that the incidences of gastrointestinal, fatigue, sleep disturbance, and allergic reaction symptoms during chemotherapy were 59.1%, 59.1%, 31.8%, and 31.8%, respectively [13]. Peripheral neuropathy is also a common adverse effect [14]. Moreover, MM patients suffer from various psychological symptoms due to the combination of the disease and the adverse effects of chemotherapy. Indeed, approximately 25% of MM patients reportedly suffer from anxiety and depression [15]. Uncertainty is a common psychological problem due to a lack of information about fear of the disease [16]. Most studies have focused on the symptoms of MM; however, fewer studies have explored the correlations between symptoms.

Network analysis could effectively explore the correlations between symptoms. Network analysis is a new method used to explore and explain the relationships between symptoms [17]. Network analysis could identify the core symptom, which is the most influencing and reinforcing symptom in the symptom network. In other words, the core symptom has the strongest strength connection with other symptoms [18]. Knowing the core symptom has the potential benefit of understanding the mechanisms involved in the onset and maintenance of the disease [Data collection

The study has shown that paper questionnaires deliver a higher response rate than e-questionnaire [31], but there has an adequate agreement between the paper and electronic questionnaires [32]. In this study, offline and online methods were used to collect data due to the COVID-19 pandemic. Two researchers checked and entered the data. Due to duplicated and omitted items, over 20% of the questionnaire responses were deleted. All the data were saved on a computer with a password and then locked in a cabinet. The data were managed by the researchers.

Offline: After obtaining the consent of the hospital and department, the patients were recruited by the medical management system. The clinical information was provided by doctors. The researchers distributed a paper version of the questionnaire to the participants. The questionnaires were anonymized and did not include identifiable information, such as name and patient ID. If the participants had any doubts about the questionnaire, they could ask the researchers. After the questionnaires were completed, the researchers immediately checked them. The researchers helped the participants fill in any missing information in the questionnaire at the end. If the participants were unwilling to complete any missing items, this study does not require the participants to fill out the questionnaires.

Online: The e-questionnaire was developed by Star Survey, an online survey platform commonly used in China. The researcher distributed the recruited information and e-questionnaire to patients via WeChat messaging. The e-questionnaire did not include any identifiable patient information. The clinical information is provided by patients according to doctor’s diagnosis.

Data analysis

Demographic and clinical information

SPSS v25.0 statistical software was used for the data analysis. Frequency and percentage were used with age, marital status, financial pressure, household unit, type of MM, and status of physical activity. Mean and standard deviation were used to analyse the duration since diagnosis, number of chemotherapy cycles, and severity of symptoms.

Network analysis

All the statistical analysis was done using R 4.1.3 and R-package qgraph. The EBICglasso and Pearson’s correlations were used to estimate the network analysis. All of the data were entered the network analysis. In the network, the symptoms are referred to “nodes”, and the “edges” are correlations between two nodes. A thicker edge represents a stronger correlation between the two symptoms.

Centrality indices

R-package bootnet was used to estimate the importance of nodes in a network and to calculate the stability of centrality indices by giving the centrality stability coefficient. The common centrality indices were strength, closeness, and betweenness. The most stable index is strength. The strength calculates the sum of the weights that the edges of a node directly connect. Closeness estimates the degree of a node indirectly connected to others nodes. This means the inverse of the weighted sum of distance between a specific node and others nodes in the network. Betweenness calculates the numbers of a node stop on the shortest path length between any two other nodes. The centrality stability (CS) coefficient should be higher than 0.25 [33]. In this study, the CS coefficient of centrality indices greater than 0.25 was explained.

Results

Demographic and clinical characteristics

Out of the 177 questionnaires collected, 7 patients’ questionnaires were excluded due to duplicated or omitted items. Thus, a total of 170 patients provided a valid questionnaire in this study. In total, 51.8% of participants were aged ≤60 years, while 48.2% were aged >60 years. The majority of participants (63.5%) felt great financial pressure. The type of IgG κ (23.5%) and IgG λ (22.9%) accounted for half of the participants. Most participants had physical activity scores of 1–3 (90.5%), while few participants had scores of 0 (2.4%) or 4 (7.1%). The demographic and clinical characteristics of the participants are shown in Table 1.

Table 1 The demographic and clinical characteristics of participants with MM (n = 170)
Full size table

The symptoms experienced by the participants

Table 2 shows the symptoms experienced by the participants. The prevalence of symptoms is 12.9–71.8%. The most prevalent and distressing symptoms were pain (71.76%), lack of energy (67.06%), and difficulty slee** (63.53%). The most severe symptoms were problems with sexual interest or activity (1.77±1.81), pain (1.48±1.29), and lack of energy (1.38±1.24). The average scores for physical and psychological symptoms were 0.78±0.59 and 0.83±0.70, respectively. The total average score of symptoms was 0.74±0.56.

Table 2 The experienced symptoms of the participants (N=170)
Full size table

Core symptom

The network of symptoms is shown in Fig. 1. The centrality stability (CS) coefficient of strength, closeness, and betweenness were 0.28, 0, and 0, respectively. The CS coefficient of closeness and betweenness were inadequate. The primary index of centrality strength showed accepted stability (Fig. 2). The CS coefficient of strength is shown in Fig. 3.

Fig. 1
figure 1

Network of symptoms. P1, difficulty concentrating; P2, pain; P3, lack of energy; P4, cough; P5, feeling nervous; P6, dry mouth; P7, nausea; P8, feeling drowsy; P9, numbness/tingling in hands/feet; P10, difficulty slee**; P11, feeling bloated; P12, problems with urination; P13, vomiting; P14, shortness of breath; P15, diarrhoea; P16, feeling sad; P17, sweats; P18, worrying; P19, problems with sexual interest or activity; P20, itching; P21, lack of appetite; P22, dizziness; P23, difficulty swallowing; P24, feeling irritable; P25, mouth sores; P26, change in the way food tastes; P27, weight loss; P28, hair loss; P29, constipation; P30, swelling of arms or legs; P31, “I don’t look like myself”; P32, changes in skin

Full size image
Fig. 2
figure 2

Estimated strength centrality. P1, difficulty concentrating; P2, pain; P3, lack of energy; P4, cough; P5, feeling nervous; P6, dry mouth; P7, nausea; P8, feeling drowsy; P9, numbness/tingling in hands/feet; P10, difficulty slee**; P11, feeling bloated; P12, problems with urination; P13, vomiting; P14, shortness of breath; P15, diarrhoea; P16, feeling sad; P17, sweats; P18, worrying; P19, problems with sexual interest or activity; P20, itching; P21, lack of appetite; P22, dizziness; P23, difficulty swallowing; P24, feeling irritable; P25, mouth sores; P26, change in the way food tastes; P27, weight loss; P28, hair loss; P29, constipation; P30, swelling of arms or legs; P31, “I don’t look like myself”; P32, changes in skin

Full size image
Fig. 3
figure 3

Centrality stability (CS) coefficient of strength using the case-drop** bootstrap method. The x-axis represents the percentage of the original sample used for every step. The y-axis represents the average of correlation between strength centrality of the primary network and strength centrality of the re-estimated network after excluding the increasing percentage of cases. The lines represent the correlation of strength. The area indicates the 95% CI

Full size image

Item P18 (worrying) showed the highest strength centrality index (Z=0.34), followed by P7 (nausea) (Z=0.26) and P13 (vomiting) (Z=0.24). The strongest edge weight was found between P7 (nausea) and P13 (vomiting) (edge weight=0.24), followed by P16 (feeling sad) and P18 (problems with sexual interest or activity) (edge weight=0.17) and P5 (feeling nervous) and P18 (worrying) (edge weight=0.17). The estimated strength centrality is shown in Fig. 2.

Discussion

This study found that the most prevalent and distressing symptom is pain. The core symptom of MM patients undergoing chemotherapy is worrying. The strongest correlation was between nausea and vomiting.

The most prevalent and distressing symptom was pain. The study reported that approximately 70% of MM patients suffer from pain [8]. Reasons for the pain in MM patients are bone disease, peripheral neuropathy, and procedural pain due to diagnostic evaluation and treatment [34, 35]. MM patients suffer pain during all disease stages, and this influences the patient’s quality of life [36]. Recently, bisphosphonates have been recommended for the treatment of bone diseases in MM patients, as they reduce skeletal-related events and control pain [37]. The analgesic ladder, vertebroplasty, and complementary and alternative medicine were also effective in controlling pain [35]. However, the study showed that the polypharmacy increased mortality, falls, adverse drug reaction, and hospitalization and readmissions [38]. Decreasing the drug use of the patients are not necessary. In addition, other non-drug therapies, including psychotherapy [39], exercise [40], and acupuncture [41], also relieve pain. Therefore, clinical staff should relieve the pain according to each patient’s personal and disease characteristics.

In this study, worrying (a core symptom of MM patients with chemotherapy) had the most effect on other symptoms. A total of 54.1% of the participants in our study experienced worrying; this prevalence was higher than in previous studies [15]. According to the Symptom Experience Mould (SEM), personal and clinical characteristics have an effect on MM patients’ illness perception, and compared to non-chemotherapy patients, worrying includes not only worry about the disease, but also about the effectiveness of the chemotherapy and financial pressure [42]. Therefore, in our study, the patients have a higher prevalence of worrying.

Based on our findings, decreasing this worrying is beneficial in relieving other symptoms and maximizing the effectiveness of interventions. Interventions for relieving worrying are as follows:

  1. (1)

    Health education—Pereira et al. reported that patients who know more information about the disease worry less [43]. Therefore, nurses could relieve this worry by improving their patients’ knowledge about the disease.

  2. (2)

    Physical activity—international physical activity guidelines state that 150 min of moderate (or 75 min of vigorous) physical activity is beneficial in maintaining optimal psychological status [44]. Similarly, the study reported that yoga and aerobic exercise could relieve worrying [45] as the patients could get more social support [46]. Indeed, patients with higher social support have a lower incidence of worrying [47].

  3. (3)

    Aromatherapy can decrease worrying [48] as essential oils can stimulate the brain, especially the amygdala in the limbic system [49], to release 5-hydroxytryptamine and dopamine through the olfactory system to maintain mental health [50]. Therefore, due to aromatherapy with an impact in physiological, it could be a good resource to use during the chemotherapy treatment of MM.

  4. (4)

    In addition, cognitive-behavioural therapy [51], attention control [52], and comprehensive interventions [53] also relieve worrying.

Hence, suitable interventions should be used to relieve worrying, while other symptoms will also be relieved, thus maximizing the effectiveness of the interventions.

Our study found that the strongest correlation was between nausea and vomiting. A study by Cherwin also had a similar finding: nausea and vomiting were the most common gastrointestinal symptom cluster [54].

Chemotherapy damages enterochromaffin cells, thus releasing serotonin to stimulate the chemoreceptor trigger zone to release neurotransmitters that trigger nausea and vomiting [55]. Recently, patients with serious nausea and vomiting had their symptoms relieved by the administration of dexamethasone and a dopamine receptor antagonist in clinical settings [56]. In addition, attention control and ginger products also relieve nausea and vomiting [57, 58]. Hence, effective interventions are available for use in MM patients undergoing chemotherapy.

Limitations

This study has some limitations. Firstly, this study included only a small sample size, and patients in China may have a lower prevalence of symptoms. Secondly, a cross-sectional design was used in this study, and we could not explain causality among different symptoms. Next, it is necessary to conduct a similar study but with a longitudinal design. Third, since this study was self-reported, the bias was increased due to the lack of objective indicators. Finally, we were able to achieve a stable network due to the strength centrality >0.25, but the closeness and betweenness centrality were extremely low, possibly due to the small sample size.

Implications for practice

Our study is important for symptom management in chemotherapy-treated multiple myeloma patients. The core symptom of chemotherapy-treated multiple myeloma is worrying. Nurses and health care teams should be focused on the core symptom in order to achieve the maximum effectiveness of the intervention. In addition, the strongest relationship was nausea and vomiting in all of symptoms. Nurses and health care teams should be managing together.

Conclusion

In MM patients with chemotherapy, the most prevalent and distressing symptom was pain. The strongest correlation was between nausea and vomiting. Worrying was a core symptom. By focusing on worrying while managing symptoms in a clinical setting, interventions would have maximal effectiveness due to a chain reaction. Understanding the correlation among symptoms is beneficial in the management of symptoms in chemotherapy-treated MM patients. However, increasing the sample size and using longitudinal studies in the future should enable better exploration of the relationship among different symptoms.