Background
Schistosomiasis is a serious parasitic disease caused by blood flukes (trematode worms) of the genus
Schistosoma [
1]. As one of the 17 neglected tropical diseases listed by the World Health Organization, it presents the greatest global burden of disease, leading to 200 million infections and threatening 800 million residents in 78 countries worldwide [
2,
3]. Schistosomiasis is caused by
Schistosoma japonicum in China. Over the past seven decades, Chinese government has made great strides toward reducing the prevalence and incidence of schistosomiasis, largely through a strategy based on chemotherapy and snail control [
4]. According to 2017 national report of schistosomiasis, 37601 cases with
S. japonicum infection and 29407 patients suffered from advanced schistosomiasis were reported [
5]. Advanced schistosomiasis is regarded as the most severe form of schistosomiasis. This debilitating condition is associated with splenomegaly, ascites, portal hypertension, liver fibrosis and cirrhosis and gastro-oesophageal variceal bleeding, leading to disability or even death [
6]. The dwarfism and colonic tumoroid proliferation sub-types are rarely found at present while ascites and splenomegaly subtypes are still common in Hubei, a province by the Yangtze River in Central China [
7].
Cases with advanced schistosomiasis are registered and managed independently because the disease results in high levels of mortality and disability as well as poor quality of life. The prognosis of patients with advanced schistosomiasis is poor. The death is usually triggered by the upper gastrointestinal tract bleeding, hepatic failure, spontaneous bacterial peritonitis and so on [
6,
7].
Our previous study has shown that the 2-year mortality of patients with advanced schistossomiasis was as high as 8% in which the 2-year all-cause death after discharge served as the primary outcome [
8]. However, the practice in prevention and cure of advanced schistosomiasis has shown that these patients did not necessarily die as results of advanced schistosomiasis or relevant complications. For example, due to the long disease duration, patients’ condition recurrent attacks and low cure rate, the negative emotions can lead to suicide when they become severe [
9,
10]. Furthermore, some patients died by accident due to lack of social support, or from other primary diseases. In our study, non-advanced schistosomiasis-specific deaths (NASDs) and advanced schistosomiasis-specific deaths (ASDs) are regarded as two competing events in case of survival analysis. NASDs, such as primary gastric cancer, pancreatic cancer and suicide, account for a sizeable proportion. The risk of NASDs also increases as age increases. The competing risk analysis must be considered when the absolute percentage of competing events is more than 10% [
11]. Failure to take the presence of competing risks into account may result in misleading conclusions in the prognostic prediction of advanced schistosomiasis [
12]. Nonetheless, there is no study addressing the incidence of ASD and NASD events in patients with advanced schistosomiasis so far. Different from the all-cause death outcome in our previous study [
8], the survival outcome in this study was evaluated by competing risk analysis to take NASDs into account. Competing risk analysis can more adequately capture the real cause-specific survivals of the patients with advanced schistosomiasis after discharge.
The post-discharge follow-up is of considerable value for patients with advanced schistosomiasis to determine whether further treatment be needed [
13,
14]. Thus, patients with advanced schistosomiasis should receive periodic follow-up after discharge, including physical and imaging examination. Proper clinical predictive tool on patients’ survival outcome, such as nomogram, could help guide follow-up and aid accurate prognostic assessment. It is well known that personalized prevention and treatment are based on the accurate prognostic evaluation, in which clinical prognostic factors need to be clearly illustrated. For example, the competing risk analysis has been widely used in cancer research and competing risk nomograms have recently been developed for cancers such as lung cancer, thyroid cancer and renal cell carcinoma [
15‐
17]. However, competing risk nomogram for predicting mortality risk of patients with advanced schistosomiasis is still lacking.
To fill this important gap of knowledge, we studied a large population-based cohort of patients with advanced schistosomiasis from Jingzhou to construct the predictive nomograms and externally validated in Huangshi cohort [
8]. Increased awareness of the nomograms, accurate prediction of the OS and ASS rates and adequate management of prognostic factors through appropriate interventions could improve the patients’ prognosis [
18]. Therefore, competing risk nomograms, which could provide accurate individualized prognosis predictive tools in clinical practice, were utilized based on the revealed prognostic factors. This means that when evaluating the prognosis of these patients after discharge, the competing risks of ASD and NASD have to be weighed against each other based on the effective nomograms [
19].
Discussion
The disease burden of schistosomiasis is mainly attributed to the advanced stage, associated with liver fibrosis and cirrhosis, ascites, portal hypertension, enlarged spleen and gastroesophageal varices, leading to disability, loss of workforce and self help ability or even death [
21‐
23]. Previous studies have clearly showed that advanced
schistosomiasis japonica is associated with high morbidity and mortality, poor self-reported quality of life and heavy disability [
7,
8]. However, the special survival analysis literature about advanced
schistosomiasis japonicum is rare. In this study, we evaluated the 2-year survival outcome for patients with advanced schistosomiasis after discharge using competing risk approach in order to evaluate the prognosis more accurately, in which death from other causes were not censored, but treated as a competing risk failure event. We have presented the cumulative incidence of advanced schistosomiasis specific death. Cumulative incidence function (CIF) is an unbiased estimate for probability of death outcome which reflects the mortality patterns actually observed [
16]. Nomograms based on a model that includes patients’ demographic and clinical characteristics could provide an accurate individualized prediction tool which is extremely useful to guide follow-up and aid accurate prognostic assessment. This is the first study to implement competing risk analysis and build nomograms for patients with advanced schistosomiasis based on Fine and Gray’s proportional sub-distribution hazard model. The internal and external validation results have shown a favorable discrimination and calibration because the C-indexes are higher than 0.70 [
24]. The nomograms are easy to use in clinical practice because the variables incorporated in the model can be easily obtained from routine clinical work and the principle is readily comprehensible.
The results have shown that older age, splenomegaly clinical classification, abnormal serum DBil, AST, ALP and positive HBsAg were significantly associated with 2-year OS rate after discharge. Meanwhile, older age, abnormal serum AST, ALP and positive HBsAg were significantly associated with 2-year ASS after discharge. Older age, which is associated with an increasing rate of comorbidities, means a significant decline in bodily functions and thus resulting in higher mortality rate [
25,
26]. Splenomegaly type patients have many complications after splenectomy, especially combined with hepatic encephalopathy and rebleeding which may also bring about higher mortality risk than ascites type patients [
27,
28]. Serum ALB level is influenced by fundamental chronic liver dysfunction and mainly reflects the liver’s protein synthetic capability associated with severity of ascites. There is a high risk of developing further complications of hepatic cirrhosis such as Hepatorenal syndrome (HRS) and spontaneous bacterial peritonitis [
29,
30]. Even though the statistic was not significant in multivariate Cox regression analysis (
P = 0.055), the further study of the ALB role in survival outcome of patients with advanced schistosomiasis was essential. Serum ALP is a hydrolytic enzyme that dephosphorylates and transphosphorylates molecules including nucleotides (adenosine triphosphate, adenosine biphosphate), pathogen-associated molecule patterns and danger-associated molecule patterns [
31]. The serum ALP level was reported as a prognostic variable which was integrated into Chinese University Prognostic Index (CUPI) system to predict survival outcome in hepatocellular carcinoma patients [
32,
33]. Previous studies also suggested that coinfection of
S. japonicum and HBV could lead to accelerated deterioration of hepatic function and higher mortality risk [
34,
35]. Abnormal serum AST also reflects the deterioration of hepatic function and could be taken as a surrogate marker for cirrhosis because of reduced plasma clearance of AST secondary to impaired function of sinusoidal cells [
36]. Previous studies have also suggested that abnormal DBil was possibly independently associated with an increased risk of hepatic fibrosis, which represented the most common event relevant with decompensating outcome which could lead to death [
37,
38]. However, the abnormal DBil is not significantly associated with ASS in our study which could be explained by its lower specificity.
It is an innovation for applying the competing risk analysis method in patients with advanced schistosomiasis after discharge. Competing risk analysis evaluates the informative nature of censoring and the occurrence rates of a particular event, which is more suitable for prognostic analysis. Misleading conclusions might be obtained due to the failure to recognize the presence of competing risks in survival analysis [
39]. Due to its rarity of advanced schistosomiasis cases, previous studies focus more on case reports [
40,
41]. The strength of our study is that this population-based cohort has a sufficiently large sample size that includes all the local patients with advanced schistosomiasis to build reliable and effective nomograms. The internal and external validation results allow us to generalize the constructed nomograms to a larger population. Moreover, since Chinese government embarked an effort to manage theses cases independently and those registered in the Advanced Schistosomiasis Cases Management System receive an RMB 5000 subsidy yearly per capita for therapy, the patients lost to follow-up was rare [
42]. The database, which is supported by government, provides relatively complete patient data including demographic, clinical and follow-up data updated annually, which bring convenience for our research.
Our study also has several limitations. First, weaknesses inherent to the our dataset include lack of some information on some possible prognosis factors that were not routinely collected such as detailed information on treatment variables, life style and four liver fibrosis indicators. It is well-known that different specific surgeries and chemotherapies, unhealthy life styles such as smoking and excessive alcohol consumption influence survival outcomes. Hyaluronic acid (HA), Laminin (LN), Collagen IV (CIV) and Procollagen III (PCIII) may influence the survival outcome as well [
43,
44]. The patients’ ultrasonography examination results of liver and spleen were also not complete. However, models without these markers of liver fibrosis and ultrasonography examination results also performed well (all C-indexes >0.70), suggesting that the good performance of our nomograms. Second, the non-advanced schistosomiasis event, as the competing event, still contains a lot of competing events. In this study, we just regarded them as a whole, which might overstate the impact of competing event. But separating them into minute events will make it difficult for analysis because of the small sample size for each event. Third, the study cohort data, which are obtained from the Hubei Province, may not necessarily reflect the prognosis of patients in other districts very well. Although this model’s performance was internally validated with bootstrap approach and externally validated using a retrospective cohort from Huangshi city, it still needs further study using prospective cohorts.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.