Path Analysis of Hospitalization Expenses of 2,164 Appendicitis Patients at County-Level Public Hospitals in Anhui Province, China

Background and objective Appendicitis resection is one of the most common surgical procedures. Hospitalization expenses are a major determinant of appendicitis treatment. This study explored the factors inuencing hospitalization expenses of appendicitis patients in Anhui province and provided a scientic basis for reasonably controlling medical expenses. Methods A multi-stage random cluster sampling method was used to collect case information on 2,164 patients who underwent appendicitis surgery at 6 county-level public hospitals in Anhui province, China. Path analysis was used to study the factors inuencing hospitalization expenses of appendicitis patients. Results The average length of stay was 5.62 ± 2.64 days, with a median of 5 days; the average hospitalization expenses were 6,109.60 ± 2,109.44 CNY, with a median of 5,511.93 CNY. The direct effect of length of stay was 0.535. Surgical grades, surgical methods, and chronic appendicitis directly affected hospitalization expenses, with direct effects of 0.149, 0.081, and -0.037; surgical costs, anesthesia costs, disease outcomes, age, acute simple appendicitis, and operative duration not only directly affected hospitalization expenses, but also indirectly affected hospitalization expenses, the total effects were 0.283, 0.045, 0.200, 0.202, -0.162, and 0.062, respectively. The total number of surgeons and assistants only indirectly affected hospitalization expenses, with indirect effect of 0.020. Conclusions The length of stay is the most important factor affecting hospitalization expenses. Based on controlling the average length of stay, combined with shortening operative duration, conducting health education, strengthening controllable factors, and other comprehensive measures can effectively reduce the economic burden of patients and hospitals. International Classication of Disease, Tenth Revision; ICD-9-CM-3: International Classication of Disease, Ninth Revision, Clinical Modication, Third Revision.


Introduction
As a common in ammatory disease of the abdominal cavity, appendicitis is a common cause of abdominal and periumbilical pain in patients in emergency surgery. Surgical resection has been the preferred treatment [1,2]. Compared with non-surgical treatment, the success rate of appendectomy is high and the length of stay (LOS) is short, which is the gold standard for the treatment of appendicitis [3].
Among the more than 200 million operations performed each year worldwide, appendectomy is one of the most common operations that cause signi cant medical expenses [4,5,6]. The US estimated the burden of surgical disease and the results showed that appendicitis resections were numerous, accounting for the majority of hospitalizations in emergency general surgery (EGS) [7,8]. Appendectomy has also been recognized as one of the most uneven clinical burdens in EGS [9]. Among EGS diseases, appendicitis is one of the seven most expensive diseases.
In low-and middle-income countries (LMIC), the incidence of appendicitis is also high [10]. As the world's most populous and largest developing country, China has limited medical and health resources. Cost control for patients undergoing appendicitis surgery is of considerable signi cance to reduce the burden of medical treatment and the increase in medical expenses [2]. According to o cial statistics, the per capita medical expenses of appendicitis patients were 7,619.64 CNY, 8,233.5 CNY, and 8,605.5 CNY, respectively, in 2016, 2017, and 2018. The number of discharges from public hospitals for appendicitis were 789,164, 803,434, and 856,609, respectively. The numbers of appendicitis discharges from county-level hospitals were 505,456, 511,057, and 550,657, respectively. County-level hospitals accept most appendicitis patients in China, and the per capita medical expenses and the number of discharges showed an increasing trend.
Most research focuses on comparing the clinical e cacy and cost-effectiveness of open appendectomy (OA) and laparoscopic appendectomy (LA), and single or multiple factors such as region, CT scan, and antibiotic treatment have a clinical effect on appendicitis patients and the in uence of economic expenses and other factors [1,11,12,13,14]. Few studies focus on the in uence of surgical factors such as operative duration (OD), time to appendectomy (TTA), and disease outcomes on hospitalization expenses of appendicitis patients. Many surgical outcome studies focus on situations that affect the majority of the population that have a high mortality rate and risk of treatment complications [15,16]. In contrast, patients with common operations such as appendicitis usually survive. Therefore, research on the clinical outcomes of appendicitis surgery uses other indicators that re ect the quality and value of medical care, such as the LOS and hospitalization expenses [17,18]. To obtain the best therapeutic value in the treatment of appendicitis depends on the expenses, so it is appropriate to pay attention to hospitalization expenses [19].
In today's health care environment with limited resources, rising health expenditures make it increasingly important to maintain medical quality while controlling costs [20,21]. In China, the rapid increase in hospitalization expenses is an issue of public concern and a problem that urgently needs to be solved in the reform of county-level public hospitals. As leaders of the rural tertiary medical and health service network and the link between urban and rural medical and health service systems in China, county-level public hospitals have become a key link in alleviating the problem of expensive medical care for the general public.
Anhui province is located in eastern China. As the end of 2018, it had a population of approximately 70 million, and the urban population accounted for approximately 32.65%. The province is one of the four major medical reform pilot provinces; it has been at the forefront of national medical reform and county-level hospital reform and has created many innovative measures to promote throughout the country and has made considerable achievements [22]. Anhui province has continuously increased its economic investment in the health care industry. In recent years, per capita hospitalization expenses have increased only slightly and the proportion of personal health expenditures in the total health care expenditures has gradually decreased.
This study included 2,164 patients with appendicitis surgery at 6 county-level hospitals in Anhui province in 2018 as research subjects and conducted a path analysis of the factors that lead to increasing hospitalization expenses to ascertain inner connection and provide references to reduce the economic burden of patients with appendicitis surgery.

Materials And Methods
Materials and methods Sample and data collection A multi-stage cluster random sampling method was used in this study. First, we randomly selected six counties based on geographic location in northern and southern Anhui: Linquan, Taihe, Mengcheng, Wuwei, Jingcounty, and Dongzhi. We then randomly chose a county-level hospital from each of the aforementioned counties as the investigation scene. The inclusion criteria were the International Classi cation of Disease, Tenth Revision (ICD-10) code K36.x02 or K35; the International Classi cation of Disease, Ninth Revision Clinical Modi cation, Third Revision (ICD-9-CM-3) code 47.0; and the discharge time was January 1 to December 31, 2018. The exclusion criteria were missing data or errors too di cult to correct and the data was obviously abnormal. According to the inclusion and exclusion criteria, data from 2,164 appendicitis surgery patients were ultimately included. This study was approved by the Ethics Committee of the First A liated Hospital of Anhui Medical University as a retrospective large data analysis study. Since LOS Y1 and hospitalization expenses Y2 were both skewed, the frequency of the original score was converted into a relative cumulative frequency (percentage grades), which was regarded as the probability of a normal distribution. Then by checking the probability in the normal distribution table, the Z value corresponding to the value was converted into a Z score. The variable data (LOS and hospitalization expenses) that failed the normal distribution test were corrected to obtain NY1 and NY2, respectively. NY1 and NY2 were considered endogenous variables, and other relevant factors were introduced into the model. SPSS version 19.0 (SPSS Inc., Chicago, IL, USA) was used for data statistics and processing, and a multiple stepwise linear regression was used to t the path model. The signi cance was set at P ≤ 0.05, in which the disordered multi-classi cation variables entered the equation model in the form of dummy variables (see Table 1).  Table 2.  Model 1&2 shows that the total number of surgeons and assistants indirectly affected hospitalization expenses only through the LOS; factors such as surgical methods, chronic appendicitis, and surgical grades directly affected hospitalization expenses; OD, anesthesia costs, acute simple appendicitis, disease outcomes, age, and surgical costs not only directly affected hospitalization expenses, but also indirectly affected hospitalization expenses through the LOS.
Effect decomposition of the path model Each effective in uencing factor was simply correlated with NY1 (see Table 4), and the correlation coe cients ri and j were obtained. It is preliminarily believed that the closer ri and j are to 1, the greater the relationship between the two. However, because of the mutually restrictive relationship between various factors, ri and j were the result of the synthesis of many factors, and the size of ri and j could not fully explain the importance of the cause to the results. To understand the importance of the comprehensive components of ri and j and the reasons for the results, we used the path diagram and path coe cient.  The effect decomposition of in uencing factors of hospitalization expenses was carried out according to the aforementioned path model. The path coe cient signi cance test was equivalent to the partial regression coe cient hypothesis test. The results are shown in Table 5

Drawing of path diagram
Based on the direct and indirect path coe cients of various variables, combined with professional knowledge, we drew a path diagram of the in uencing factors on the hospitalization expenses of appendicitis surgery patients (see Figure 1), which visually shows the direct and indirect in uencing factors and their mutual effects of appendicitis patients.

Discussion
Many scholars have used t-test, chi-squared test, multiple regression analysis, and other methods to explore the in uencing factors of hospitalization expenses and achieved certain results [23,24,25]. However, the in uencing factors of hospitalization expenses are complex, and there are often correlations between various factors. Path analysis is an extension of multiple regression. It is a structural equation model that does not contain any latent variables [26]. We used path coe cients and standardized direct, indirect, and total effects to accurately assess the impact of each factor on the dependent variable. Path analysis can analyze the direct and indirect effects of various factors on dependent variables, reveal the relationship between variables, and compensate for the regression model's de ciencies. Therefore, this study adopted this method.

The in uence of LOS on hospitalization expenses
As one of the important indicators re ecting the progress of the disease, LOS is the greatest factor in uencing the hospitalization expenses of appendicitis patients. The direct effect is 0.535, indicating that the longer LOS, the higher hospitalization expenses. Therefore, to ensure medical quality, how to effectively shorten LOS has become a key issue in reducing hospitalization expenses. The average LOS should be shorted by formulating diagnosis and treatment standards or clinical pathways, optimizing the diagnosis and treatment process, reducing the number of days of hospitalization before surgery and the number of ine cient and ineffective hospitalization days for patients. In addition, patients who meet the standards should be promptly referred to township health centers or community health service centers for rehabilitation, thereby effectively decreasing the postoperative hospital stay. Based on the perspective of health economics, decreasing the number of days in the hospital can expedite hospital bed turnover and increase the medical resources utilization rate; the number of patients admitted to the hospital per unit time will increase, and the total income will increase, which not only reduces average expenses, but also patients' economic burden.
The in uence of surgical grades and methods on hospitalization expenses The two variables are factors that only had direct effect on hospitalization expenses and the total effect was positive. The result showed an increasing trend as the surgical grades increase, which was consistent with the results of the descriptive analysis. The increase in hospitalization expenses may be related to the increased di culty of the operation and the increase in medical equipment consumables.
Surgical methods positively correlated with hospitalization expenses, indicating that the average hospitalization expenses of patients undergoing LA were higher than the average hospitalization expenses of patients undergoing OA, which was consistent with the results of the descriptive analysis. LA has been recognized as a surgical technique that can optimize patient safety, quality, and cost [27,28]. The bene ts of LA include reduced surgical site infection, less postoperative pain and need for analgesia, and shorter LOS. Long-term bene ts include reduced adhesion formation, earlier resumption of normal activities, and cost savings due to earlier resumption of work [27,29,30]. Relevant studies have shown that LA is not fully accepted in some places. This may be because LA requires more manpower and material resources and higher hospital expenses, although it may save money if the patient can resume work faster [30,31,32,33,34]. Prior studies reported a higher incidence of patients with simple appendicitis [35], similar to the results of this study. Analysis of age groups shows that the average LOS of patients does not gradually increase with age. The average LOS of 18 to 33 years old patients is lower than that of other age groups. This may be due to the low coexistence characteristics of young patients, morbidity, and better tolerance to surgical pressure (see Table 6). The in uence of surgical costs, anesthesia costs, disease outcomes, age, and OD on hospitalization expenses These ve variables not only directly affect hospitalization expenses, but also indirectly affect hospitalization expenses through LOS, and the total effect is positive. This demonstrates that hospitalization expenses will increase as surgical and anesthesia costs increase. A series of comprehensive service expenses, medical equipment consumables, and examination and diagnosis expenses incurred during surgery directly led to the increase in hospitalization expenses. To reduce hospitalization expenses, surgeons should improve the level of medical technology, reduce the loss of medical devices, and control unreasonable surgical costs.
The average hospitalization expenses of patients who are converted to other are higher than that of patients who are converted to improved, and the average hospitalization expenses of improved patients are higher than that of cured patients. Patients with better outcomes have a relatively short LOS, consume less medical equipment and medical supplies, and have lower hospitalization expenses. The direct effect is greater than the indirect effect, indicating that the outcome mainly directly affects hospitalization expenses.
Age is uncontrollable and the third major factor affecting hospitalization expenses. Age is positively correlated with hospitalization expenses. This may be due to aging and damage to human organs, as well as the gradual weakening of physical functions with age, various complications, and accompanying diseases, resulting in a relatively long LOS and higher hospitalization expenses. In this regard, the public should increase their awareness of appendicitis and carry out health education, such as avoiding irregular eating, strenuous exercise after meals, spicy and irritating food, and indigestible food such as raw, cold, and hard food. The general public should be instructed to develop good dietary and living habits to fundamentally prevent the occurrence of diseases and achieve early detection and treatment to reduce the economic burden.
Hospitalization expenses will increase as the OD increases. This is not di cult to understand, mainly because of the increase in medical equipment, consumables, and medical supplies.
The in uence of acute simple appendicitis and chronic appendicitis on hospitalization expenses The total effect of these two variables is negative. Compared with patients with acute comorbid appendicitis, patients with acute simple appendicitis have lower average hospitalization expenses. Patients with more serious comorbidities and those with comorbidities have more complicated conditions than patients without comorbidities and will consume more sanitary supplies and medicines during the treatment process, which in turn increases the expenses of medicines. This is inconsistent with the results of the descriptive analysis and may be caused by this study's small cohort. Studies have shown that for patients undergoing acute appendicitis with abscesses and other complications, conservative treatment is better than immediate surgical treatment at reducing morbidity, hospitalization expenses, and loss of quality of life [36].
Compared with patients with acute comorbid appendicitis, patients with chronic appendicitis have lower average hospitalization expenses. This is consistent with the results of the descriptive analysis. Patients with acute appendicitis have more complicated and urgent conditions than those with chronic appendicitis. During treatment, they will consume more sanitary supplies and medicines, which in turn increases the costs of medicines and inspections and ultimately leads to an increase in hospitalization expenses.
The in uence of the total number of surgeons and assistants on hospitalization expenses The total number of surgeons and assistants only indirectly affects the hospitalization expenses through LOS and is positively correlated with hospitalization expenses. This is inconsistent with the aforementioned descriptive analysis results. It is possible that the path analysis results are more scienti c and accurate, so further analysis and research are needed.

The in uence of other factors on hospitalization expenses
In this study, the TTA had no signi cant in uence on hospitalization expenses. However, previous studies have shown that hospital work e ciency during weekends is lower than during workdays [37,38,39]. Compared with patients admitted on weekdays, patients admitted on weekends have a higher mortality rate and less reliable diagnostic imaging results [40,41]. Although the incidence of many medical emergencies does not change between workdays and weekends, there are several factors that can negatively affect patient outcomes during workdays and weekends, such as a reduction in the number of medical staff and access to laboratories and diagnostics [42,43]. In addition, inexperienced medical staff and imaging workers can take off work in shifts after work, which may become a problem that reduces the e ciency of medical care on weekends and weekday evenings [44]. Surgery should be scheduled during the day on weekdays, which will help improve the utilization and effectiveness of hospitals' medical resources and reduce medical expenses [45,46]. Prior studies demonstrated that the readmission rate after appendectomy is estimated to be between 1% and 19% [47,48,49], which showed that improving the e ciency of diagnosis and timely management, especially the value of nursing care, is bene cial to hospitals [50]. Primary medical and health care providers should guide patients to develop good dietary and living habits to reduce the readmission rate and effectively decrease hospitalization expenses.
The path coe cient can be used to calculate the effect of factors that have not been studied but actually have an impact, that is, the remaining path coe cient, which can prompt whether it is necessary to further search for other in uencing factors. Pe = 0.699, which is slightly higher than the direct path coe cient of hospitalization days, indicating that there are other factors that have a greater impact on hospitalization costs. This demonstrates that there are still some factors that have a greater impact on hospitalization costs, such as drug expenses and payment methods. There were many missing data in this research, which made the residual path coe cient higher, which was the main limitation of this study. In addition, the cost was related to the economic development grades. This research only selected Anhui province, which has a middle national economic development grade. In the future, the scope of research should be expanded to nd national factors in comprehensively developed and undeveloped areas. This was another limitation of this study that should be further researched.

Conclusions
This study analyzed the direct and indirect effects of hospitalization expenses for patients undergoing appendicitis surgery. The research results provide evidence of a better understanding of how factors such as surgical methods affect hospitalization expenses. Hospitals should focus on bed turnover e ciency to reduce the number of bed days to reduce pressure and consider rational use of clinical pathways and day surgery to control hospitalization expenses. This study provided recommendations for policies and measures to reduce hospitalization costs for patients undergoing appendicitis surgery to promote the reform of county-level public hospitals.

Declarations
Ethics approval and consent to participate The Ethics Committee of the First A liated Hospital of Anhui Medical University approved this project (procedural no. Quick-PJ 2020-11-07).

Consent for publication
We have obtained the agency's publishing consent.

Figure 1
Path diagram of the in uencing factors on the hospitalization expenses of appendicitis surgery patients. Figure 1