Trends of hospitalisation for patients with liver cirrhosis in Ningxia, China: a cross-sectional study

Abstract

Objective

To evaluate the changing trends of hospitalisation for patients with liver cirrhosis between 2015 and 2019 by using hospitalisation summary records in Ningxia Hui Autonomous Region (NHAR), China.

Design

A cross-sectional study.

Setting

Hospitalisation summary records between 1 January 2015 and 31 December 2019 from 28 top-ranking hospitals in NHAR were extracted and rigorously analysed.

Participants

During the study period, hospitalisation records referring to liver cirrhosis were included. Records with missing data were excluded. A total of 16 566 patients with liver cirrhosis were included in this study.

Outcome measures

International Classification of Diseases codes, tenth version (ICD-10) and text-diagnoses were used to identify hospitalisation records referring to liver cirrhosis.

Results

Between 2015 and 2019, hospitalisation rates for liver cirrhosis declined from 8.38% to 5.57%. Chronic viral hepatitis accounted for almost 70% of all liver cirrhosis admissions; the remaining 30% of patients were admitted due to non-viral hepatitis cirrhosis (28.06%) and alcoholic cirrhosis (2.05%). The male-to-female hospitalisation rate ratio for liver cirrhosis was 2.57. The hospitalisation rate for workers with hepatitis cirrhosis was significantly higher than farmers (hospitalisation rate ratio (RR)=1.06, 95% CI 1.01 to 1.15, p<0.001); this was also the case for alcoholic cirrhosis (RR=5.23, 95% CI 3.34 to 8.20). However, the hospitalisation rate for workers with non-viral hepatitis cirrhosis was significantly lower than for farmers (RR=5.23, 95% CI 3.34 to 8.20, p<0.001). The hospitalisation rate increased in patients over the age of 30 years and reached a peak at the age of 45–50 years.

Conclusions

The hospitalisation rate for liver cirrhosis has declined over recent years and chronic viral hepatitis remains the major cause of liver cirrhosis in NHAR. Hospitalisation summary records can efficiently reflect the local changing trends of hospitalisation for liver cirrhosis and represent an efficient strategy for the surveillance of chronic disease.

Strengths and limitations of this study.

• This study enrolled a large sample size using electronic medical records in Ningxia Hui Autonomous Region, which provided a significant opportunity to reliably analyse hospitalisation rate trends of liver cirrhosis in the region.

• This study developed a rigorous process of data extraction, cleaning, processing, case ascertainment and privacy protection, which provided a model for the analysis of electronic medical data.

• There were few studies on the prevalence and risk factors of liver cirrhosis in less developed areas of western China and this study could fill in some knowledge gap.

• Hospitalisation summary records may contain measurement errors, including incomplete or inaccurate information, as well as processing errors, which may have an impact on the results of the current study.

Introduction

Liver cirrhosis is one of the leading causes of death worldwide and represents an advanced stage of chronic liver disease.¹ Over the last few decades, the number of newly diagnosed cases of liver cirrhosis has continued to increase globally, although many public health initiatives have been implemented to counteract this trend.² Chronic viral hepatitis and alcohol consumption are considered to be the major global aetiologies of liver cirrhosis.³

The epidemiology and aetiologies of liver cirrhosis in each region or country are known to vary based on a number of different risk factors that remain poorly understood.⁴ For instance, in the USA, most European countries and Japan, hepatitis C virus and alcohol are the most common causes, whereas hepatitis B virus-related liver cirrhosis is predominant in China.^5–10 In addition, with rapid socioeconomic changes and hepatitis vaccination, the epidemic pattern of liver cirrhosis, along with its aetiologies, may have changed; however, specific information relating to this is scarce, particularly in low-income and middle-income countries.¹

Population-based disease surveillance systems (PBDS) are usually considered to be an ideal data source for information relating to the prevalence of diseases in a defined population.¹¹ However, in addition to the use of PBDS for some major diseases, such as population-based cancer registries, there are few active disease surveillance systems for most chronic diseases.

One emerging resource for the surveillance of the risk factors and conditions of chronic disease is the data that can be acquired from electronic health records (EHRs).¹² In China, the use of data from EHRs, such as medical records, which features data relating to health reimbursement, has been regarded as an efficient strategy for the surveillance of chronic disease over recent years.^13–15 However, only a limited number of studies have used hospitalisation summary records (HSRs) to monitor chronic disease.

In the present study, we developed a rigorous procedure and used data acquired from 28 hospitals in Ningxia Hui Autonomous Region (NHAR), an underdeveloped region in Northwest China to evaluate the changes in hospitalisation rate for patients with liver cirrhosis between 2015 and 2019. The study involved more than 2.37 million hospitalised patients. With such a large sample, our analysis provided a significant opportunity to reliably analyse hospitalisation rate trends according to major types of liver cirrhosis and demographic groups.

Methods

Data sources

Data were obtained from 2015 to 2019 by accessing the HSRs of 28 hospitals in the NHAR. Hospitals in China are divided into three grades, primary (town or community-level), secondary (county-level) and tertiary (city-level and above). This hospital ranking relates to infrastructure, level of services and the quality and safety of care. Secondary and tertiary hospitals represent the highest levels of regional medical care, and generally receive patients referred from primary hospitals. The hospitals selected in the present study included 9 tertiary and 19 secondary hospitals.

The data acquired from the hospital HSRs were electronically submitted to the NHAR Municipal Health Bureau through a centralised health information system, according to administrative requirements of the Ministry of Health. The medical information acquired from the HSRs included basic demographics, dates of admission and discharge, eight discharge diagnoses in Chinese and corresponding International Classification of Diseases codes, tenth version (ICD-10, one principal and seven supplementary diagnoses), treatments (mainly surgical information including date, coding, anaesthetist and surgeon), the outcome of hospitalisation (survival status, drug allergy and hospitalisation infection) and financial costs.

Data security and confidentiality

We established strict procedures to ensure the security of data during data extraction, storage, processing and statistical analysis. To protect patient privacy, we hid personally identifiable information at the personal level, including name, identity (ID) number, phone number and home address. We then generated a unique, encrypted hexadecimal identifier for each patient as a label to further ensure that duplicate diagnostic records are eliminated.

Study patients

The procedure for identifying study patients is shown in figure 1. We extracted 3 473 703 hospitalisation records from 28 hospitals between 1 January 2013 and 31 December 2019 using an annual time frame based on the date of admission from the health information system in the NHAR. Records with missing data (ID number and name (N=7188)) were excluded. Because the unit of analysis was an individual patient rather than a hospitalisation record, when the hospitalisation rate was calculated, we selected the first hospitalisation record for each patient due to similar diseases during the study period as a study record (excluding readmission, N=535 844). To reduce the impact of prevalent cases, records from 1 January 2013 to 31 December 2014 (N=560 129) were excluded. Finally, a total of 2 370 542 hospitalised patients were included in the study.

Figure 1.

The procedure for capturing patients with liver cirrhosis from the health information system in Ningxia Hui Autonomous Region. ^aRecords were exported based on the date of admission. ^bRecords were identified as liver cirrhosis if they met any of the following criteria: ^aICD-10 codes included K70.2, K70.3 and K74; ^btext-diagnoses referring to liver cirrhosis occurring in any of the eight listed diagnostic codes. ^cICD-10 codes included B15 to B19 and Z22.5. ^dCirrhosis other than viral cirrhosis and alcoholic cirrhosis. ^eICD-10 codes included K70.2 and K70.3. ICD-10, International Classification of Diseases codes, tenth version.

On the basis of the hospitalised patient population, we used ICD-10 codes (K70.2, K70.3 and K74) and text-diagnoses to identify hospitalisation records referring to liver cirrhosis occurring in any of the eight listed diagnostic codes. A total of 16 566 records with liver cirrhosis were identified.

To explore the aetiology of liver cirrhosis, we distinguished viral hepatitis cirrhosis from non-viral hepatitis cirrhosis by examining whether patients had a diagnosis of viral hepatitis (ICD-10 codes: B15-B19, Z22.5) in any of the eight diagnoses. Finally, we identified 11 577 patients with viral hepatitis cirrhosis and 4649 with non-viral hepatitis cirrhosis. In addition, we also identified 340 hospitalised patients with non-viral hepatitis and alcoholic cirrhosis (K70.2 and K70.3).

Statistical analysis

All of the selected demographic characteristics were subjected to calculation of descriptive statistics. Age was summarised as median (IQR) and other variables were displayed as percentages. The hospitalisation rates for patients with liver cirrhosis were calculated by gender, occupation, medical institution level, source of patients and year and the χ² test was adopted to compare the hospitalisation rates among subgroups. To address these changes, we used a Poisson regression model to estimate rate ratio (RR) and 95% CI. To fit models, we established a data set of dependent variables (including the numbers of patients with viral hepatitis cirrhosis, non-viral hepatitis cirrhosis and alcoholic cirrhosis (numerators), the number of hospitalised patients (denominator) and independent variables (including gender and occupation)). Using the hospitalisation rate for women and farmers as a reference, we estimated RRs and 95% CIs for men and other occupations. An RR>1 with a p value<0.05 indicated a statistically significant change in the rate of hospitalisation for liver cirrhosis.

All statistical analyses were performed using Stata/SE V.14.0 for Windows (Stata Corporation, College Station, Texas, USA).

Patient and public involvement

None.

Results

Demographic characteristics and hospitalisation rates of liver cirrhosis

The demographic characteristics of the study patients and hospitalisation rates for liver cirrhosis are shown in table 1. The 16 566 patients with liver cirrhosis in this study, 11 340 (68.45%) were men and 5226 (31.55%) were women. The median age was 53 years and approximately 40% of patients with liver cirrhosis were farmers. In addition, almost 80% of patients were diagnosed and treated in tertiary hospitals, and 93.45% of patients were native to the NHAR.

Table 1.

Selected demographic characteristics and hospitalisation rates for liver cirrhosis by gender, age, occupation and medical institution level among hospitalised patients in 28 top-ranking hospitals in Ningxia Hui Autonomous Region, China, 2015–2019

Variable	Hospitalised patients	Liver cirrhosis		P value
		Patients	Rates*
	n (%)	n (%)	% (95% CI)
Gender				<0.001
Male	1 176 348 (45.78)	11 340 (68.45)	9.64 (9.46 to 9.81)
Female	1 393 219 (54.22)	5 226 (31.55)	3.75 (3.65 to 3.85)
Age, year				<0.001
Median (quartile)	49 (27–65)	53 (45–64)	–
Occupation				<0.001
Farmer	775 540 (30.18)	6625 (39.99)	8.54 (8.34 to 8.75)
Worker	85 498 (3.33)	737 (4.45)	8.62 (8.01 to 9.26)
Office clerk	156 986 (6.11)	809 (4.88)	5.15 (7.81 to 5.52)
Retired or unemployed	768 478 (29.91)	4633 (27.97)	6.03 (5.86 to 6.20)
Others	783 065 (30.47)	3762 (22.71)	4.80 (4.65 to 4.96)
Medical institution level				<0.001
Secondary	1 052 863 (40.97)	3467 (20.93)	3.29 (3.18 to 3.40)
Tertiary	1 516 704 (59.03)	13 099 (79.07)	8.64 (8.49 to 8.79)
Source of patients†				<0.001
Ningxia	2 405 033 (93.45)	15 082 (91.06)	6.27 (6.17 to 6.37)
Other provinces or autonomous region	168 576 (6.55)	1481 (9.04)	8.79 (8.35 to 9.24)
Year				<0.001
2015	473 371 (18.42)	3967 (23.95)	8.38 (8.12 to 8.64)
2016	505 221 (19.66)	3217 (19.42)	6.37 (6.15 to 6.59)
2017	518 410 (20.17)	3212 (19.39)	6.20 (5.98 to 6.41)
2018	526 345 (20.48)	3130 (18.89)	5.95 (5.74 to 6.16)
2019	546 220 (21.26)	3040 (18.35)	5.57 (5.37 to 5.77)
Total	2 569 567 (100.00)	16 566 (100.00)	6.45 (6.35 to 6.55)	–

*Hospitalisation rates of patients with liver cirrhosis among hospitalised patients in 28 hospitals in Ningxia between 2015 and 2019.

†Analysis were limited to individuals who had complete residence information in the hospitalisation summary records.

When considering all hospitalisation rates, the specific proportion of hospitalisation for liver cirrhosis was 6.45‰. The hospitalisation rates for men, workers, patients from tertiary medical institutions, and other provinces or autonomous regions, were significantly higher than for women and other occupations, and patients from secondary medical institutions and those who were native to the NHAR (p<0.001). In addition, the hospitalisation rates for liver cirrhosis declined from 8.38‰ to 5.57‰ between 2015 and 2019 (p<0.001).

Distribution of aetiologies

The distribution of aetiologies for liver cirrhosis cases is presented in table 2. Almost 70% of the hospitalised patients with liver cirrhosis were due to viral hepatitis cirrhosis; the other 30% were due to non-viral hepatitis cirrhosis (28.06%) and alcoholic cirrhosis (2.05%). The rates of hospitalisation were significantly higher in men than in women for all three types of liver cirrhosis (p<0.001).

Table 2.

Aetiology of the enrolled patients with liver cirrhosis by gender among hospitalised patients in 28 top-ranking hospitals in Ningxia Hui Autonomous Region, China, 2015–2019

Aetiologies of cirrhosis	Male and female		Male		Female		P value
	Cases	Hospitalisation rate	Cases	Hospitalisation rate	Cases	Hospitalisation rate
	n (%)	% (95% CI)	n (%)	% (95% CI)	n (%)	% (95% CI)
Viral hepatitis cirrhosis	11 577 (69.88)	4.51 (4.42 to 4.59)	8420 (74.25)	7.16 (7.01 to 7.31)	3157 (60.41)	2.27 (2.19 to 2.35)	<0.001
Non-viral hepatitis cirrhosis	4649 (28.06)	1.81 (1.76 to 1.86)	2587 (22.81)	2.20 (2.12 to 2.29)	2062 (39.46)	1.48 (1.42 to 1.55)	<0.001
Alcoholic cirrhosis	340 (2.05)	0.13 (0.12 to 0.15)	333 (2.94)	0.28 (0.25 to 0.32)	7 (0.13)	0.00 (0.00 to 0.01)	<0.001
Total	16 566 (100.00)	6.45 (6.35 to 6.55)	11 340 (100.00)	9.64 (9.46 to 9.82)	5226 (100.00)	3.75 (3.65 to 3.85)	<0.001

Hospitalisation rate ratios and 95% CIs for gender and occupation by types of liver cirrhosis

Table 3 shows the comparisons of gender and occupation with regards to the rate of hospitalisation according to the three types of liver cirrhosis. The gender ratio was 2.25 (95% CI 2.16 to 2.35; p<0.001) for viral hepatitis cirrhosis, 1.06 (95% CI 1.00 to 1.12; p<0.001) for non-viral hepatitis cirrhosis and 40.17 (95% CI 19.00 to 84.91, p<0.001) for alcoholic cirrhosis. Compared with farmers, the hospitalisation rate for workers was significantly higher for viral hepatitis cirrhosis (RR=1.06; 95% CI 0.97 to 1.15; p<0.001) and alcoholic cirrhosis (RR=5.23; 95% CI 3.34 to 8.20), but significantly lower for non-viral hepatitis cirrhosis (RR=5.23; 95% CI 3.34 to 8.20; p<0.001). In other occupation groups, the hospitalisation rates decreased, except for alcoholic cirrhosis.

Table 3.

Hospitalisation rate ratios and 95% CIs* for year, age and gender by types of liver cirrhosis Ningxia Hui Autonomous Region, China, 2015–2019

Variables	Viral hepatitis cirrhosis	Non-viral hepatitis cirrhosis	Alcoholic cirrhosis
Gender
Female	Reference	Reference	Reference
Male	2.25 (2.16 to 2.35)	1.06 (1.00 to 1.12)	40.17 (19.00 to 84.91)
	p<0.001	p<0.05	p<0.001
Occupation
Farmer	Reference	Reference	Reference
Worker	1.06 (1.01 to 1.15)	0.73 (0.61 to 0.87)	5.23 (3.34 to 8.20)
Office clerk	0.59 (0.54 to 0.64)	0.54 (0.46 to 0.63)	3.80 (2.52 to 5.74)
Retired or unemployed	0.62 (0.59 to 0.65)	0.91 (0.85 to 0.98)	2.52 (1.83 to 3.48)
Others	0.51 (0.49 to 0.53)	0.68 (0.63 to 0.74)	1.68 (1.19 to 2.36)
	p<0.001	p<0.001	p<0.001
Year
2015	Reference	Reference	Reference
2016	0.74 (0.70 to 0.78)	0.81 (0.74 to 0.88)	0.71 (0.51 to 0.98)
2017	0.73 (0.69 to 0.77)	0.76 (0.70 to 0.83)	0.73 (0.53 to 1.02)
2018	0.73 (0.69 to 0.77)	0.65 (0.59 to 0.71)	0.80 (0.58 to 1.10)
2019	0.66 (0.62 to 0.69)	0.69 (0.63 to 0.75)	0.65 (0.47 to 0.91)
	P for trend <0.001	P for trend <0.001	P for trend <0.001

P values for linear trend were obtained from modelling the continuous form of year variable.

*Hospitalisation rate ratios and 95% CIs in the table were estimated by Poisson regression models.

Age-specific hospitalisation rates for liver cirrhosis by gender

Finally, we compared the age and gender distribution of the hospitalisation rates for liver cirrhosis in the NHAR, 2015–2019 show in figure 2. The hospitalisation rate was relatively low in individuals less than 30 years-of-age, and increased in those over 30 years-of-age, both of men and women, reaching a peak at ages of 45−50 years. The ratio for male-to-female hospitalisation rate for liver cirrhosis was 2.57.

Figure 2.

Age and gender distribution of the hospitalisation rates for liver cirrhosis in Ningxia Hui Autonomous Region, China, 2015–2019.

Discussion

Accurate data referring to the prevalence and risk factors of liver cirrhosis are essential for evidence-based healthcare planning and resource allocation; however, up-to-date data are scarce in the NHAR, China. Using the HSRs from 28 top-ranking hospitals in the NHAR, the present study adopted a uniform and standardised scientific methodology to provide a significant opportunity to estimate the burden of liver cirrhosis and its related characteristics in NHAR, China.

Between 2015 and 2019, the hospitalisation rates for liver cirrhosis declined by approximately 3% and chronic viral hepatitis remained the major cause of liver cirrhosis, accounting for almost 70% of all liver cirrhosis admissions. This finding was consistent with previous studies reported for other regions in China.^{5 6}Because of the extremely high prevalence of hepatitis B in the Chinese population, the high hospitalisation rate for viral hepatitis cirrhosis may be associated predominantly with hepatitis B rather than with other types of viral hepatitis.¹⁶ With hepatitis B vaccination and improvements in the efficacy of antiviral therapy for hepatitis over recent decades, the hospitalisation rates for liver cirrhosis have decreased.

The distribution of liver cirrhosis aetiologies in NHAR was somewhat different from that in eastern China.⁵ For example, liver cirrhosis caused by chronic viral hepatitis accounted for 60% of all liver cirrhosis admissions in a study conducted in Beijing but accounted for 70% in our present study. Furthermore, alcoholic cirrhosis accounted for 13% and 2% of hospitalisations in Beijing and the NHAR, respectively. This may be related to the development of regional society and economy. In China, there is an imbalance in economic development when compared between the East and the West; the East is much more developed than the West. Furthermore, the use of alcohol in developed regions is far greater than in less developed areas.¹⁷ The prevalence of viral hepatitis in less developed areas was higher than in developed regions due to poor hygiene conditions.¹⁸ Strengthening the health management of people infected with hepatitis B is the key to reduce the incidence of liver cirrhosis, it may be effective to incorporate health management of patients with hepatitis B into basic public health services.

The higher rate of hospitalisation for liver cirrhosis observed in men in the present study might be due, at least in part, to the higher frequencies of alcohol use and viral hepatitis infection among the male population. The hospitalisation rate was shown to be higher in both men and women aged over 30 years and reached a peak at 45−50 years; this is consistent with the career development curve of most subjects. Work and family commitments have led to many bad living habits, thus leading to liver cirrhosis.¹ Paying attention to diet, reasonable nutrition, abstinence from drinking, strengthening labour healthcare and avoiding all types of chronic chemical poisoning, are all positive measures for prevention.

In addition, we observed a higher rate of hospitalisation for viral hepatitis cirrhosis in manual labours such as farmers and workers, who usually had a limited education level and healthcare provision. Many patients with non-severe disease, such as non-fatal liver diseases, never go to see a doctor or attend hospital until the condition becomes critical.¹⁵ Higher levels of alcoholic cirrhosis were detected in office clerks and those who are retired; these patients have a relatively good economic status and a higher frequency of alcohol use. It makes all the difference that the focus of health education for different groups is very important. Considering liver cirrhosis management, it could be implemented in the focus populations, for example, low-income groups and retired elderly people.

The outline of China’s 13th Five-Year Plan (2016–2020) and Healthy China 2030 plan pointed out that the integration, sharing, mining and application of big data in different health fields should be vigorously promoted. In 2017, the National Health and Family Planning Commission proposed to accelerate the information construction of national major chronic disease monitoring and management. In this study, we established the value of HSR in assessing trends in hospitalisation for cirrhosis, suggesting that HSR can effectively reflect the local burden of cirrhosis, which will provide new perspectives for the reform and development of China’s chronic disease surveillance strategy.

Due to the large sample size in this study, and the combined representativeness of the entire NHAR population, HSR is certainly of great value for studying temporal trends in disease. Despite there being quality control measures HSR data still contains measurement errors, including incomplete or inaccurate information, as well as processing errors. Therefore, misclassification of the disease may have occurred, especially in the cirrhosis subtype. To avoid potential misclassification, we included three main types of cirrhosis in our analysis, instead of numerous subtypes of cirrhosis. The main types of cirrhosis are cause-specific, so it can help identify trends and associated causes. Because HSR was limited to hospitalised patients, non-hospitalised patients with cirrhosis could not be analysed. However, fatal diseases such as cirrhosis often required multidisciplinary and comprehensive treatment in high-level hospitals, which were rarely provided by primary hospitals and there were little outpatients.

Conclusions

Herein, we conducted a large data-descriptive analysis of hospitalisation trends and gender ratio of liver cirrhosis in the NHAR. The hospitalisation rate of liver cirrhosis has declined over recent years and chronic viral hepatitis remains the major cause of liver cirrhosis. It may be effective to incorporate health management of patients with hepatitis B into basic public health services to reduce the incidence of liver cirrhosis. Hospitalisation summary records can efficiently reflect the local changing trends of hospitalisation of liver cirrhosis, which represents an efficient strategy for chronic disease surveillance.

Data availability statement

Data are available upon reasonable request.

Ethics statements

Patient consent for publication

Not applicable.

Ethics approval

Data were anonymised before the authors accessed them for the purpose of this study. The research protocols used in the present study were approved by the Institutional Review Board of the People's Hospital of NHAR (2020-ZDYF-018), NHAR, China.