ORIGINAL RESEARCH

How Serious is the Economic Burden of DiabetesMellitus in Hainan Province? A Study Basedon ‘‘System of Health Account 2011’’

Yuanyuan Dong . Chunping Liu . Peng Zhou . Yalan Zhu .

Qingcheng Tang . Siyu Wang . Xin Wang

Received: September 6, 2019 / Published online: October 18, 2019� The Author(s) 2019

ABSTRACT

Introduction: The treatment of diabetesrequires extensive use of healthcare resources,resulting in high medical costs, which in turnplaces a heavy economic burden on society,patients and their families.Methods: A multi-stage stratified random sam-pling method was used to investigate 283medical and health institutions in Hainan

Province. The total medical expenses relating todiabetes in Hainan Province in 2016, institu-tional flow directions, the composition of ser-vice functions and the distribution of thehealthcare costs to beneficiaries were analyzedbased on the System of Health Account 2011.The STATA version 12.0 statistical softwarepackage was used to collate operation data, andSPSS software was used to carry out regressionanalysis on the factors affecting hospitalizationcosts.Results: In 2016, the total medical expenses forthe treatment of diabetes in Hainan Provincewas 242.17 billion renminbi (RMB), of which81.95% was spent in high-level hospitals and14.71% was spent in medical institutions pro-viding primary care. There was little differencebetween outpatient and hospitalization expen-ses (53.01 and 46.99%, respectively). Hospital-ization accounted for 77.62% of the expenses ofmedical institutions providing primary care.Older patients were found to spend more onmedical and drug expenditure.Conclusion: The economic burden of health-care expenses for the treatment of diabetes inHainan Province is massive, and patient treat-ment is concentrated in large hospitals. It isnecessary to inform patients to focus more onmedical institutions that provide primary care,adjust the proportion of medical insurancereimbursement, control the cost of hospitaliza-tion and strengthen the healthcare manage-ment of middle-aged and elderly diabetic

Yuanyuan Dong and Chunping Liu are co-first authors.

Enhanced Digital Features To view enhanced digitalfeatures for this article go to https://doi.org/10.6084/m9.figshare.9948575.

Y. Dong � Y. Zhu � X. Wang (&)College of the Humanities and Social Sciences,China Medical University, Shenyang, People’sRepublic of Chinae-mail: wxinsmile@qq.com

C. Liu � P. ZhouAdministration School, Hainan Medical University,Haikou, People’s Republic of China

Q. TangMedical Information School, Hainan MedicalUniversity, Haikou, People’s Republic of China

S. WangClinical Medical Science, China Medical University,Shenyang, People’s Republic of China

X. WangSchool of Public Health, Xinjiang MedicalUniversity, Urumqi, People’s Republic of China

Diabetes Ther (2019) 10:2251–2263

https://doi.org/10.1007/s13300-019-00712-0

patients. Only in this way can costs be reducedand the economic burden be eased.

Keywords: Diabetes; Medical expenses; Systemof Health Account 2011

Key Summary Points

Why carry out this study?

The treatment of diabetes consumesconsiderable healthcare resources,resulting in high medical costs, whichplaces a heavy economic burden onsociety, patients and their families.

The aim of this research was to study, onthe scale of medical expenses for diabeticpatients in Hainan Province in 2016, thedistribution of the expenses of thebeneficiaries and the flow direction ofinstitutions, to provide a reference forhealth departments to formulate healthpolicies and allocate medical resourcesrationally.

What was learned from the study?

The medical burden of diabetestreatment in Hainan Province is heavy,and patient treatment is concentrated inlarge hospitals. It is necessary to informpatients to focus more on medicalinstitutions that provide primary care,adjust the proportion of medicalinsurance reimbursement, control thecost of hospitalization and strengthen thehealthcare management of middle-agedand elderly diabetic patients.

The medical expenses of diabetestreatment mainly flow to high-levelmedical institutions, and medicalinstitutions providing primary care fail tofunction fully in their role of‘‘gatekeeper.’’ Early prevention and earlytreatment of diabetes mellitus will likelybe the key to reduce the burden of medicalexpenses for diabetes mellitus in thefuture.

INTRODUCTION

Diabetes mellitus is a group of metabolic dis-eases characterized by hyperglycemia [1]. Withthe rapid economic developments and theabrupt transition of lifestyles and society inrecent years, morbidity and mortality linked todiabetes are increasing rapidly worldwide [2],with the number of diabetic patients increasingyear by year. According to a recent survey,around 415 million adults globally have dia-betes, of whom approximately 110 million arein China [3]. China has become the second-largest country with diabetes after India.Updated data show that the estimated overallprevalence of diabetes and prediabetes in Chi-nese adults is 10.9 and 35.7%, respectively [4].

Diabetes, accompanied by multiple compli-cations, cannot be cured and requires life-longmedication. Patients with diabetes are fourfoldmore likely to suffer from cerebrovascular dis-ease and heart disease than those without dia-betes, and the mortality of a diabetic personwith hypertension is sevenfold higher than thatof a person with uncomplicated hypertension[5]. The treatment of diabetes requires extensiveuse of medical resources, resulting in highmedical costs. These costs are a heavy economicburden on society, patients and their familiesalike and have a serious impact on socialdevelopment at both the individual and popu-lation level. Studies have shown that diabetes isbeginning to shift to younger age groups, suchas teenagers, with this trend growing in recentyears. This trend towards rejuvenation is aninvisible burden to the family [6]. With theadvent of complications, there will also beconflicts between work and treatment, whichwill also negatively affect the economic burdenof the family and society. It is therefore evidentthat diabetes has become a public health prob-lem that cannot be ignored.

In the study reported here, we have evalu-ated the economic burden of diabetes patientsseeking medical treatment in Hainan Province.We chose Hainan Province because it provides asignificant reference point for such studies inChina, with average levels of both economicand medical development in comparison to

2252 Diabetes Ther (2019) 10:2251–2263

other provinces. We also established a compa-rable database for other provinces, which willallow us to perform similar studies in the future.It must be noted that the study of the diseaseburden is still in its infancy in Hainan Province,with only a few large-scale and high-qualitystudies on this subject in Hainan Province [7].To our knowledge, no research report on thecost of diabetes treatment in Hainan Provincehas ever been published. In the present study,we studied the scale of medical expenses fordiabetic patients in Hainan Province in 2016,the distribution of the expenses of the benefi-ciaries and the flow direction of institutions,with the aim to provide a reference for healthdepartments to formulate health policies andallocate medical resources rationally.

The System of Health Account 2011 (SHA2011) is a well-known health cost accountingsystem worldwide that was formulated by theWorld Health Organization (WHO), the Orga-nization for Economic Cooperation and Devel-opment (OECD) and the European UnionStatistical Office (EUROSTAT) [8]. The applica-tion of this new accounting system to a study ofmedical expenses in Hainan Province can bringthese expenses better in line with internationalstandards and improve the international anddomestic comparability of these expenses inHainan Province. It can help contribute tomodifications of health policies and the alloca-tion of medical resources in Hainan Province.

This overall purposes of this study was toevaluate the economic burden of diabetespatients in seeking medical treatment by cal-culating the medical cost of diabetes in HainanProvince and to propose countermeasures toimprove the medical condition of patients andrelieve the economic pressure.

METHODS

Data Source

All of the data were derived from the annualreport of health statistics of Hainan Province in2016 and the annual report of health finance ofHainan Province in 2016. The health statisticsyearbook and health financial annual report

were provided by the Hainan Health CareerDevelopment Center. The case data came from283 medical institutions and public healthinstitutions in Hainan Province through asampling survey.

Data Sample

The study used a multistage stratified randomsampling method. The first stage was to selectthe sample cities and counties, taking intoaccount the level of economic development,medical services and medical informationmanagement system in Hainan Province, aswell as other factors. We chose a prefecture-levelcity, Haikou, and three counties, includingWenchang (considered to be a county-levelcity), Chengmai and Dingan. The second stagewas to select survey areas. For this purpose, weextracted two districts from Haikou—Qiong-shan district and Longhua district—and threetowns and three villages from each of the threecounties. In the third stage, we selected medicalinstitutions in the selected areas according tothe level and type of medical institutions. Atotal of 283 health institutions were selected,including ten in Hainan Province, eight inHaikou City, 45 in Wenchang City (county), 39in Qiongshan District, 54 in Longhua District,73 in Chengmai County and 54 in DinganCounty. The data on outpatient and inpatientcases collected from the sample survey institu-tions were classified on the basis of the medicaland health system in Hainan Province. Once allinvalid or incorrect data had been removed, theavailable data encompassed a total of 2,529,946samples. The basic information includedpatient’s sex, age, disease diagnosis, medicalexpenses, type of medical insurance payment,type of medical institutions, among other fac-tors. Patients with endocrine, nutritional andmetabolic diseases in Hainan Province in 2016were included in the study. The types of dia-betes (type 1 diabetes, type 2 diabetes, malnu-tritional diabetes) and their complications werestudied. Patient information was normalizedaccording to the ‘‘International classification ofdiseases, tenth revision’’ (ICD-10; https://www.who.int/classifications/icd/en/). The main

Diabetes Ther (2019) 10:2251–2263 2253

factors influencing hospitalization medicalexpenses of 12,171 diabetic patients were sta-tistically analyzed.

Statistical Methods

The accounting principles and methods adop-ted in the study are according to the revisededition of ‘‘A system of health accounts 2011’’[8] and are further explained in the followingsections.

Accounting of Medical ExpensesThe medical expenses (CME) covered in thisstudy include treatment income and basicexpenditure subsidy, and the outpatient andinpatient medical expenses are included in theservice function. The formula is as follows:

CME ¼X

K¼1

ðCKINCþCKSUBÞ

ðKmeans different types of medical institutionsÞð1Þ

where CKINC represents the treatment income ofvarious medical institutions, and CKSUB repre-sents the basic expenditure subsidy of variousmedical institutions.

To calculate treatment income, we used thefollowing formula:

CINC ¼ CTINC 1� ap�a

� �ð2Þ

where CTINC represents the actual total incomeof different medical institutions, a representsthe total income of sample institutions, ap rep-resents the total income of prevention servicesin the sample and (1 - ap)/a represents theproportion of treatment income of sampleinstitutions to the total income of sampleinstitutions after excluding prevention services.

The actual treatment income is expressed byCIINC, where ai represents the treatment incomeof the case in the sample, and ai/(a - ap) rep-resents the proportion of the treatment incomeof the case in the sample after excluding theprevention service income. The formula is asfollows:

CIINC ¼Xm

i

CINCai�

a� ap� �

ð3Þ

The basic expenditure subsidy is allocatedaccording to the proportion of service in thesample. CISUB represents the total actual basicexpenditure subsidy of all medical institutions,CTSUB represents the total basic expendituresubsidy of various medical institutions, CPSUB

represents the basic expenditure subsidy ofpreventive services and S represents the totalamount of service in the sample and caseservices in the sample. The amount isexpressed in Si, and the amount of preventiveservices in the sample is expressed as Sp. Theformula is as follows:

CISUB ¼Xm

i

CTSUB � CPSUBð Þsi�

s� sp� �� �

ð4Þ

Analysis of Factors InfluencingHospitalization Expenses for DiabetesMultiple linear regression modeling was used toanalyze the hospitalization expenses and factorsinfluencing diabetes, including age, sex, hospi-talization time, operation, insurance paymenttype, medical institution type, among others, inorder to determine the most important factors.In the multiple linear regression models, thehospitalization expenses of diabetic patientswere consided to be dependent variables, therelative variable standard was 0.05 and the cri-terion for eliminating variables was 0.10. Thestatistical analysis was conducted using SPSSversion 20.0 software (IBM Corp., Armonk, NY,USA).

Compliance with Ethics Guidelines

All procedures performed in studies involvinghuman participants were in accordance withthe ethical standards of the Ethics Committee ofHainan Medical University and with the 1964Helsinki declaration and its later amendmentsor comparable ethical standards. All informedconsent forms and the data used were recordedand kept in accordance with guidelines by theHealth Economics Association of Hainan Pro-vince. The informed consent forms and the data

2254 Diabetes Ther (2019) 10:2251–2263

used were provided to the Ethics Committee ofHainan Medical University, which granted per-mission for the study.

RESULTS

Medical Expenses for Managementof Diabetes Patients and InfluencingFactors

The medical expenses due to the managementof diabetes patients in Hainan Province can beclassified as outpatient expenses and hospital-ization expenses, with the outpatient expensescalculated to be 121.84 billion renminbi (RMB)and the hospitalization expenses calculated tobe 120.87 billion RMB. The proportion of out-patient expenses and hospitalization expensesto the total medical costs of managing diabeteswere 50.04 and 49.96%, respectively. In 2016,the costs of drugs for the management of dia-betes mellitus amounted to 94.35 billion RMB,accounting for 38.96% of the total treatmentexpenses for diabetes, of which 58.90% was inan outpatient setting and 41.10% was in ahospital setting.

The results of multiple stepwise regressionanalysis showed that F = 1089.99, p B 0.001,thereby establishing the regression equation.From the standardized regression coefficient,the top three factors influencing hospitalization

expenditure were the length of stay, type ofinsurance and age (p\0.05) (Table 1).

The Institutional Flow of MedicalExpenses

According to the analysis of the flow directionof diabetes medical expenses in Hainan Pro-vince, the majority of patients’ medical expen-ses flowed to the higher-level hospitals (generalhospitals and hospitals of traditional Chinesemedicine); this amounted to 198.47 billionRMB, accounting for 81.95% of patients’ totalmedical expenses. The proportion of drugcharges in patients’ costs in hospitals was47.54%, and the proportion of inspectioncharges was 24.01% (the proportion of drug andinspection charges in general hospitals was46.83 and 21.51%, respectively, and that intraditional Chinese hospitals was 31.65 and29.56%, respectively). A total of 35.62 billionRMB flowed to basic medical institutions (in-cluding township hospitals, community healthservice centers, village clinics and individualclinics), accounting for 14.71% of total patients’costs. In primary hospitals, the proportion ofdrug charges in patients’ costs was 18.09%, andthe proportion of inspection charges was 8.27%.Only a small proportion of the medical expen-ses was spent in outpatient service institutionsand public health institutions (3.27 and 0.07%,respectively; Fig. 1).

Table 1 Factors influencing medical expenses for the management of diabetes patients analyzed by multiple linearregression

Model Unstandardized coefficients Standardized coefficient

B Standard error B t statistic p

Constant 11,588.107 1370.894 8.453 \ 0.001

Length of stay 802.009 13.069 0.467 61.367 \ 0.001

Type of insurance 3307.969 86.271 0.301 38.344 \ 0.001

Age 96.727 3.083 0.235 31.372 \ 0.001

Operation - 16,321.107 646.305 - 0.181 - 25.253 \ 0.001

Hospital level 2246.043 127.329 0.155 17.64 \ 0.001

Type of hospital - 176.498 20.147 - 0.076 - 8.761 \ 0.001

Sex 650.588 128.961 0.038 5.045 \ 0.001

Diabetes Ther (2019) 10:2251–2263 2255

Composition of Service Functionsof Different Medical Institutions

According to our analysis of the composition ofmedical expenses for the management of dia-betes mellitus in different types of medicalinstitutions in Hainan Province, outpatient

expenses accounted for 50.04% of medical costsand inpatient expenses accounted for 49.96%.The total, outpatient and inpatient proportionsof medical expenses for the management ofdiabetes mellitus and total costs according todifferent types of medical institutions in HainanProvince are shown in Fig. 2 and Table 2,respectively.

Fig. 1 The proportion of medical expenses according to medical institution type

Fig. 2 Composition of service functions of different medical institutions

2256 Diabetes Ther (2019) 10:2251–2263

Regarding outpatient expenses of the differ-ent institutions managing diabetes mellitus inHainan Province, the proportion of primarymedical and health institutions and outpatientinstitutions was determined to be 13.11%. Atotal of 86.83% of outpatient expenses flowedto major hospitals. Based on the analysis of theinpatient expenses of diabetes, general hospitalsand traditional Chinese hospitals accounted for37.78 and 39.30% of total outpatient expenses,respectively.

Distribution of Diabetes TreatmentExpenses in Different Age Groups

To further observe the role of age in expensesdue to diabetes mellitus management and

hospitalization, we divided diabetic patientsaged\ 75 years into five age groups. We foundthat 71.88% of the people aged[45 years spentthe largest proportion of their medical expenseson diabetes mellitus management; this wasespecially true for those aged 60–74 years forwhom diabetes management accounted for29.99% of total medical expenses (Table 3). Thetreatment cost of young and middle-aged dia-betic patients aged 30–44 was also very high,accounting for 14.5% of total medical expenses.Of all drug expenses in Hainan Province,75.56% were consumed by diabetic patientsaged[ 45 years. The drug expenses of middle-aged and elderly diabetic patients aged [60 years exceeded half of the total expendituresfor diabetic drugs. In 2016, the compositionratio of the drug costs of diabetes mellitus inHainan Province was similar to that of thetreatment expenses (Table 3).

DISCUSSION

The incidence of diabetes in China hasincreased rapidly in recent years, reaching 9.7%of the general population in 2017 [9]. There areabout 136 million diabetes patients worldwide,and diabetes has become a serious public healthproblem. There are many risk factors for dia-betes, and the disease is a life-long condition.With disease progression, patients may develop

Table 2 Allocation of medical expenses according to different levels of medical institution

Institution Total Outpatient setting Inpatient setting

BillionRMB

Proportion(%)

BillionRMB

Proportion(%)

BillionRMB

Proportion(%)

Hospital 198.4699 81.95 105.2187 86.83 93.2512 77.08

General hospital 136.9160 56.54 91.2111 75.27 45.7049 37.78

Traditional Chinese medicine

hospital

61.5539 25.42 14.0077 11.56 47.5463 39.30

Basic medical institution 35.6186 14.71 7.9716 6.58 27.6471 22.85

Outpatient department 7.9142 3.27 7.9142 6.53 – –

Public health institution 0.1679 0.07 0.0792 0.07 0.0887 0.07

RMB Renminbi

Table 3 Age distribution of medical expenses for diabetes

Age group(years)

Medical expensesratio (%)

Drug expensesratio (%)

0–14 4.74 3.6

15–29 8.88 7.91

30–44 14.5 11.51

45–59 23.56 24.06

60–74 29.99 31.13

C 75 18.33 20.37

Diabetes Ther (2019) 10:2251–2263 2257

any number of a variety of severe complica-tions, such as blindness, kidney failure andgangrene. In China, 73.2% of patients withdiabetes suffer from chronic complications, andthe treatment costs of these complicationsaccount for[80% of the total medical costs ofpatients with diabetes. Among these chroniccomplications, cardiovascular disease, cere-brovascular disease and kidney disease are themain causes of the disability and death ofpatients with diabetes. The medical expensesgenerated by this disease are also the mostimportant part of the total expenses in China[10]. The medical expense burden of diabetespatients is so heavy that in 2016 the totalmedical expenses added up to 100 billion RMB,accounting for 5% of the medical expenses forall chronic non-communicable diseases. Theout-of-pocket costs accounted for 30% of thetotal expense. Diabetes mellitus places tremen-dous financial pressure on Chinese society andon the individual patient’s family. The risk of adiabetic patient’s family falling into poverty ishigh [11].

In our study, the medical expenses formanaging diabetes in Hainan Province addedup to 242.17 billion RMB, accounting for 28.3%of the total medical expenses for managingendocrine, nutritional and metabolic diseases(855.71 billion RMB). This suggests that gov-ernment and health institutions should paymore attention to the financial burden of dia-betes and implement additional measures tocontrol the medical expenses. We also foundthat in terms of diabetes management, themedical and drug expenses for outpatients arehigher than those for inpatients. One reason isthat patients prefer outpatient treatment andtherefore spend more on drug expenses thaninpatients because this incurable chronic dis-ease requires long-term medication. Drugs,diagnosis and treatment costs are long-termburdens for patients with chronic diseases [12].Cost ratios between people with and withoutdiabetes have been reported to range from 1.5to 4.4, which is different from cost ratios inother diseases [13–15]. Jacobs et al. found thelargest differences in health expenses camefrom prescribed medication in pharmacies (costratio diabetes:no diabetes: 2.2) [16].

We found that the age and sex of the patient,number of days hospitalized, type of operation,type of medical insurance payment, type ofmedical institution, level of medical institutionand hospitalization expenses had a significantimpact on the total medical expense, in accor-dance with results reported previously [17]. Themedical cost of surgical patients is significantlyhigher than that of other patients, which maybe a direct result of the severity of the diseaseand the higher cost of treatment. Medicalinsurance has been found to have the potentialto promote an increase in medical expenses,with health insurance impacting on a patient’sability to pay and a doctor’s behavior regardingdiagnosis and treatment [18]. In our study, wereached the same conclusion. One explanationis that patients reimbursed by medical insur-ance can afford more treatment than self-paidpatients. The former prefer high-tech, highlyeffective therapies and expensive drugs whenchoosing treatment methods and means, sotheir costs are likewise higher than those of self-paid patients. The longer the hospital stay percapita, the higher the cost of treatment. Com-pared with non-surgical patients, surgicalpatients stay in the hospital longer, and medicalexpenses naturally increase. The medicalexpenses of patients with diabetes mellitus andassociated conditions are considerably higherthan those of other patients, which may bedirectly related to the severity of the disease andthe cost of treatment [19].

In terms of overall cost, hospital expenditureoccupies the dominant position, with drug andinspection charges in hospitals being higherthan those of other institutions. The drug andinspection charges of primary institutions are\20% of total medical expenditure, possibly dueto patients’ trust in hospitals being higher thanthat in primary healthcare institutions. Due tothe serious shortage of medical technical per-sonnel at the primary care level, patients cannotget access to a superior hospital for further studyfor a long time, and it would appear that theprofessional level of these personnel cannot betimely and effectively improved. At the sametime, the serious shortage of funds in medicalinstitutions providing primary care is a com-mon social phenomenon, Therefore, the

2258 Diabetes Ther (2019) 10:2251–2263

technical staff and equipment of primary insti-tutions cannot meet the needs of patients, sothey cannot gain the trust of ‘‘patients withminor diseases.’’ As a result, the outpatient ser-vice and inpatient visits of grassroot institutionsare seriously insufficient, which leads to variousdrug and inspection charges being fewer andlower [20, 21]. The acceleration of urbanizationand the lag in providing basic health facilities tomeet demand means that high-quality medicalresources are concentrated in large generalhospitals [22]. Consequently, Chinese patientsprefer hospitals for better treatments, a prefer-ence that is related to the functional orienta-tion, diagnosis and treatment capacity of suchinstitutions. Overall, the role of primaryhealthcare institutions in the treatment of dia-betes needs to be further strengthened [23].According to the policy of graded diagnosis andtreatment in China, especially the prevention ofdiabetes, the management of diabetes patients,the prognosis of disease, the rehabilitation ofprimary medical and health institutions and therole of medical and health institutions at alllevels can be improved. The present state is oneof the reasons why the economic burden ofdiabetes is so heavy.

Regarding the service function of medicalexpenses of diabetes, there is a large differencebetween Hainan Province and the whole coun-try, with the costs of hospitalization for diabetesin Hainan Province being about ten percentagepoints higher than the overall national level.This may be due to the high prevalence of dia-betes mellitus in Hainan Province and the long-term hospitalization required. An analysis of thedifferent types of medical institutions revealedthat most expenses are spent in outpatient costsand drugs. In Ulrich et al.’s study, diabetespatients with insulin monotherapy or with acombination of insulin and oral antidiabeticdrugs had significantly elevated costs comparedwith individuals without diabetes (3.98 vs. 2.58,respectively) [24].

Diabetes mellitus is common among theelderly: the higher the age, the higher themedical expense [25]. Our research showed thatthe hospitalization expenses of patients aged\75 years old increased with age. However, in thefinal analysis, it is not age itself that was found

to cause elevated medical expenses; rather, thedecline of the functions of the human immunesystem and organs play a role. A decline oforgan functions translates into more basic dis-eases; consequently, elderly diabetic patientsmay have one or more complications and needmore care as the disease progresses, resulting inhigher costs than patients in other age groups[26]. In recent years, however, there has been atrend for diabetic patients to be younger,attributed in part to the large changes in life-style and the living environment, greaterworking pressure, disordered dietary structureand irregular lifestyle that have appeared in thelast decades. In comparison to older patients,patients in the younger age groups have shortdisease onset time, relatively single-treatmentindicators, fewer complications and lowertreatment costs. Thus, the results of our agedistribution analysis gives a clear indicationthat people in the age group associated withhigh expenses should pay great attention to theprevention of diabetes.

Suggestion to Policy-Makers

Controlling Hospitalization Expensesand Improving Medical InsuranceStudies on hospitalization costs for the man-agement of diabetes mellitus and other diseaseshave shown that the number of hospitalizationdays are a positive correlation factor affectinghospitalization costs and that, consequently,limiting hospitalization days is an effective wayto control medical costs [27]. Diabetic patientscan reduce hospitalization rates by regulatingtheir diet to control blood sugar, taking medi-cine on time and undergoing regular examina-tions to reduce complications.

The medical insurance management depart-ment should improve the level of medicalinsurance supervision and management, moni-tor the expenditure of hospitalization expensesfor patients with different types of medicalinsurance, ensuring it is kept within a reason-able range, ensure the rational use of the fundsavailable and regulate the unreasonableincrease in hospitalization expenses. Reim-bursement to the outpatient is still too low,

Diabetes Ther (2019) 10:2251–2263 2259

resulting in large out-of-pocket expense.Therefore, the medical insurance managementdepartment should also pay more attention tothis reimbursement ratio and make someadjustments.

Using Hierarchical Diagnosisand the Treatment System to Control MedicalExpenseMedical expenses for diabetes in Hainan Pro-vince were mainly in high-level medical insti-tutions, and there is little patient flow toprimary healthcare institutions. Therefore, toreduce the economic burden of disease, it isnecessary to improve both the role of primaryhealthcare institutions in the treatment of dia-betes and the healthcare system so that patientsreceive good advice in choosing a primaryhealthcare institution [28]. Healthcare man-agement institutions should strengthen thevocational skills training of medical staff inprimary institutions, dispatch doctors fromhigher hospitals to attend clinics at the primarycare level and improve the ability of healthcareworkers at the grassroots level to managechronic diseases similar to diabetes. The gov-ernment should attach importance to the con-struction of basic facilities and support andencourage primary care medical institutions tocarry out public health services, with the aim toincrease the rate of patients visiting primaryhealthcare institutions.

Strengthen Prevention Measures and Curbthe Prevalence of Diabetes

In the early stage of diabetes, strict control ofblood glucose is an effective measure to reduceand delay microvascular complications [29, 30].The American Diabetes Association published‘‘Diabetes Medical Standards’’, which recom-mends that adult diabetes screening shouldstart at age 45 years [31]. Basic medical institu-tions provide comprehensive preventive carefor elderly patients with diabetes, includinglectures on diabetes that provide patients withknowledge of the disease that will enable themto better interpret and communicate their con-dition and interpret the problems they may

encounter, and conducting research on thepatient’s daily diet, exercise habits, blood sugarmonitoring, and drug use. Specific preventiveand healthcare measures can be formulatedaccording to each patient’s condition in termsof lifestyle changes, exercise and diet regimens,massage, among others. Nursing staff shouldregularly conduct telephone return visits andfamily return visits and establish online com-munication groups to solve the problemsencountered by patients in real-time. In thisway, patients can be encouraged to maintaingood living habits in terms of diet, exercise andlifestyle [32].

Study Limitations

Diabetes on its own is associated with highhealthcare costs, but complications of the eye,kidney and cardiovascular and nervous systemsmay also arise, and these significantly increasethe medical expenses. In this study we adoptedSHA 2011, an internationally accepted health-care accounting method, to study only the costsassociated with the primary diagnosis andtreatment of diabetes; the complications ofdiabetes were not included in the calculation.Future studies will explore the costs related todiabetes and complications. We only researchedthe direct financial burden of medical expensesfor treating diabetic patients; any indirect costburdens were not considered. In future studies,we will continue to explore the indirect burdenaccounting method so as to more accuratelyassess the economic pressures placed on dia-betes patients.

CONCLUSION

This study on the medical expense of diabetespatients in Hainan province revealed that theeconomic burden of health expenses of diabetespatients is relatively heavy. Our study empha-sizes the importance of reducing the economicburden of Hainan residents and of identifyingthe main factors which influence the medicalexpenses. We suggest promoting the hierarchi-cal diagnosis and treatment system, improvingthe reimbursement ratio of medical insurance,

2260 Diabetes Ther (2019) 10:2251–2263

perfecting drug policies and strengthening pre-vention policies.

ACKNOWLEDGEMENTS

The authors express thanks to China NationalHealth Development Research Center for theirhelp with SHA 2011 analysis. We also gratefullyacknowledge the support of Hainan provincialhealth commission in providing data. We alsowould like to thank all the participants in thisstudy.

Funding. This work was supported byNational Natural Science Foundation of China(71673299) and the Social Science Fund ProjectPlanning of Shenyang City (18014). The RapidService Fee was funded by the authors.

Authorship. All named authors meet theInternational Committee of Medical JournalEditors (ICMJE) criteria for authorship for thisarticle, take responsibility for the integrity ofthe work as a whole, and have given theirapproval for this version to be published.

Authorship Contributions. XW designedthe study, helped in implementing the projectand reviewed the manuscript. YD and CL werethe main drafters of the manuscript. YD wasresponsible for the revision of the manuscript.PZ and YZ did some of the expeditionaryresearch. QT and SW contributed to the coor-dination and manuscript editing. All authorsparticipated in the analysis and discussion ofthe research, under the leadership and instruc-tions of XW. All authors read and approved thefinal manuscript.

Disclosures. Yuanyuan Dong, Chun-ping Liu, Peng Zhou, Yalan Zhu, QingchengTang, Siyu Wang and Xin Wang have nothingto disclose.

Compliance with Ethics Guidelines. Allprocedures performed in studies involvinghuman participants were in accordance with

the ethical standards of the Ethics Committee ofHainan Medical University and with the 1964Helsinki declaration and its later amendmentsor comparable ethical standards. All informedconsent forms and the data used were recordedand kept in accordance with guidelines by theHealth Economics Association of Hainan Pro-vince. The informed consent forms and the dataused were provided to the Ethics Committee ofHainan Medical University, which granted per-mission for the study.

Data Availability. The datasets generatedduring and/or analysed during the currentstudy are available from the correspondingauthor on reasonable request.

Open Access. This article is distributedunder the terms of the Creative CommonsAttribution-NonCommercial 4.0 InternationalLicense (http://creativecommons.org/licenses/by-nc/4.0/), which permits any noncommercialuse, distribution, and reproduction in anymedium, provided you give appropriate creditto the original author(s) and the source, providea link to the Creative Commons license, andindicate if changes were made.

REFERENCES

1. Dall TM, Yang W, Gillespie K, et al. The economicburden of elevated blood glucose levels in 2017:diagnosed and undiagnosed diabetes, gestationaldiabetes, and prediabetes. Diabetes Care.2019;42(10):dc181226. https://doi.org/10.2337/dc18-1226.

2. Jing Z, Chu J, Imam Syeda Z, et al. Catastrophichealth expenditure among type 2 diabetes mellituspatients: a province-wide study in Shandong,China. J Diabetes Investig. 2019;10(2):283–9.

3. International Diabetes Federation. IDF diabetesatlas, 7th edition. Online version of IDF diabetesatlas. 2015. http://www.diabetesatlas.org.

4. Wang L, Gao P, Zhang M, et al. Prevalence andethnic pattern of diabetes and prediabetes in Chinain 2013. JAMA. 2017;317(24):2515–23.

Diabetes Ther (2019) 10:2251–2263 2261

5. King H, Aubert RE, Herman WH. Global burden ofdiabetes, 1995–2025: prevalence, numerical esti-mates, and projections. Diabetes Care.1998;21(9):1414–31.

6. Zhichao Z. Study on age distribution characteristicsand related factors of diabetes mellitus in youngpatients. Med Forum. 2018;10:1304–5.

7. Qixin D, Hui W, Huimei M, Xuanxuan W, Shang-cheng Z, Xiang Z. Analysis of influencing factors ofeconomic burden of diseases among elderlypatients with common chronic diseases in HaikouCity. Med Soc. 2019;6:39–42.

8. OECD, Eurostat, WHO. A system of health accounts2011—revised edition. Paris: OECD Publishing;2011.

9. Yang W, Lu J, Weng J, et al. Prevalence of diabetesamong men and women in China. N Engl J Med.2010;362(12):1090–101.

10. He X, Zhang Y, Ruan Z, Li L, Wu J. The prevalenceand related direct medical costs of chronic com-plications among patients with type 2 diabetes inChina. Value Health. 2018;21:S37.

11. Yuhui Z, QuanW, Peipei C, Feng G, Xiufeng W. Themedical costs and financing burden of diabetesmellitus in China. Chin Health Econ.2017;36(4):17–9.

12. Zongjie L, Wenfang C, Wenwen Z, Caijun Y, Shi-min Y, Yu F. Medication analysis of hypertensionand diabetes mellitus in primary medical institu-tions in Shaanxi Province in 2016. Chin PharmMatters. 2017;12:1497–505.

13. Ng CS, Lee JY, Toh MP, Ko Y. Cost-of-illness studiesof diabetes mellitus: a systematic review. DiabetesRes Clin Pract. 2014;105(2):151–63.

14. Rosella L, Lebenbaum M, Fitzpatrick T, et al. Impactof diabetes on healthcare costs in a population-based cohort: a cost analysis. Diabetic Med.2016;33(3):395–403.

15. Sortsoe C, Green A, Jensen PB, Emneus M. Societalcosts of diabetes mellitus in Denmark. Diabet Med.2016;33(7):877–85.

16. Jacobs E, Hoyer A, Brinks R, Icks A, Kuß O, Rath-mann W. Healthcare costs of type 2 diabetes inGermany. Diabet Med. 2017;34(6):855–61.

17. Cao P, Wang K, Zhang H, Zhao R, Li C. Factorsinfluencing the hospitalization costs of patientswith type 2 diabetes. Asia Pacif J Public Health.2015;27[Suppl 2]:55S–60S.

18. Yang C, Huang Z, Sun K, Hu Y, Bao X. Comparingthe economic burden of type 2 diabetes mellituspatients with and without medical insurance: across-sectional study in China. Med Sci Monit.2018;24:3098–102.

19. Bao X, Yang C, Fang K, Shi M, Yu G, Hu Y. Hospi-talization costs and complications in hospitalizedpatients with type 2 diabetes mellitus in Beijing,China. J Diabetes. 2017;9(4):405–11.

20. Xianwen L. Consideration on related issues ofgrading treatment model. Health Econ Res.2015;3:18–20.

21. Zhu S. Building talented workforce teams is the keyto strengthened primary care. Chin J General Prac-titioners. 2016;15(8):599–600.

22. Wang Y, Sun L, Hou J. Hierarchical medical systembased on big data and Mobile internet: a newstrategic choice in health care. JMIR Med Inform.2017;5(3):e22. https://doi.org/10.2196/medinform.6799.

23. Hongwei T. Community outpatient expenses andmanagement of key population with type 2 dia-betes mellitus Physical analysis. China MaternalChild Health Study. 2017;28[S1]:682–3.

24. Ulrich S, Holle R, Wacker M, et al. Cost burden oftype 2 diabetes in Germany: results from the pop-ulation-based KORA studies. BMJ Open.2016;6(11):e012527.

25. Yang Y, Goh SY, Tan SB, et al. The burden of dia-betes mellitus in elderly patients from an Asiantertiary hospital. Eur J Internal Med.2012;23(1):e1–4.

26. Jiarong D, Lu L, Wantao D. Analysis of influencingfactors of hospitalization costs for diabetic patientsin a hospital. Chongqing Med. 2018;47(29):3773–5.

27. Yaling D, Zhong W. Multivariate regression analysisof influencing factors on hospitalized intensivetreatment of type 2 diabetes mellitus. China MedRecords. 2016;17(8):49–52.

28. Lingzhi Z. Experience and revelation from pilottrials of first and two-way referral in urban com-munities in China. Chongqing Med.2010;39(2):250–1.

29. Zoungas S, De Galan BE, Ninomiya T, et al. Com-bined effects of routine blood pressure lowering andintensive glucose control on macrovascular andmicrovascular outcomes in patients with type 2diabetes: new results from the ADVANCE trial.Diabetes Care. 2009;32(11):2068–74.

2262 Diabetes Ther (2019) 10:2251–2263

30. ADVANCE Collaborative Group, Patel A, MacMa-hon S, et al. Intensive blood glucose control andvascular outcomes in patients with type 2 diabetes.N Engl J Med. 2008;358(24):2560–72.

31. [No authors listed]. Standards of medical care indiabetes—2016: summary of revisions. DiabetesCare. 2016;39[Suppl 1]:S4–5.

32. Yiwei G, Pu Z, Jiuhong M, et al. Effects of integratedintervention of community traditional Chinesemedicine on diabetes patients. Shanxi J Med.2019;48(15):1852–3.

Diabetes Ther (2019) 10:2251–2263 2263