Evaluating Disease Management Program Effectiveness

Disease management (DM) program evaluations are somewhat limited in scope because of typically small sample sizes comprising important subsets of the treated population. Identifying subsets of the data that have differing results from the aggregate of the whole program can lend insight into where, when, and how the program achieves its results. Additionally, there is a very limited set of classical tools available for the smaller sample sizes typically encountered in DM. Without readily available standard error and confidence interval (CI) calculations, the analyst may be fooled by specious details.A method called the ‘bootstrap’ is introduced as a suitable technique for allowing DM program evaluators to use a broader array of quantities of interest and to extend inferences to the population based on results achieved in the program. The bootstrap uses the power of modern computers to generate many random samples from a given data set, allowing the use of repeated samples’ statistic (e.g. mean, proportion, and median). Using a congestive heart failure (CHF) program as an example, the bootstrap technique is used to extend a DM program evaluation beyond questions addressed using classical statistical inference: (i) how much of a median cost decrease can be expected as a result of the program?; (ii) did the program impact the highest and lowest costing members equally; and (iii) how much of a decrease in the proportion of patients experiencing a hospitalization can be expected as a result of the program?The potential advantages of the bootstrap technique in DM program evaluation were clearly illustrated using this small CHF program example. A more robust understanding of program impact is possible when more tools and methods are available to the evaluator. This is particularly the case in DM, which is inherently biased in case-mix (e.g. strive to enroll sickest first), often has skewed distributions or outliers, and may suffer from small sample sizes.The bootstrap technique creates distributions that allow for a more accurate method of drawing statistical inferences of a population. Moreover, since classical statistical inference techniques were designed specifically for parametric statistics (i.e. assuming a normal distribution), the bootstrap can be used for measures that have no convenient statistical formulae. Additionally, CIs can be defined around this statistic, making it a viable option for evaluating DM program effectiveness.     

The Indiana Chronic Disease Management Program

The Indiana Chronic Disease Management Program (ICDMP) is intended to improve the quality and cost-effectiveness of care for Medicaid members with congestive heart failure (chronic heart failure), diabetes, asthma, and other conditions. The ICDMP is being assembled by Indiana Medicaid primarily from state and local resources and has seven components: (1) identification of eligible participants to create regional registries, (2) risk stratification of eligible participants, (3) nurse care management for high-risk participants, (4) telephonic intervention for all participants, (5) an Internet-based information system, (6) quality improvement collaboratives for primary care practices, and (7) program evaluation. The evaluation involves a randomized controlled trial in two inner-city group practices, as well as a statewide observational design. This article describes the ICDMP, highlights challenges, and discusses approaches to its evaluation.     

Impact of a Migraine Disease Management Program

Objective: To assess the impact of a comprehensive migraine disease management program, as measured by humanistic outcomes measures, conducted in a managed care setting. Design: A prospective comparative study comprised of an intervention and a control group to evaluate the impact of the disease management initiative. Setting: Independent Practice Association (IPA)-type managed care organization. Study participants and main outcomes measures: Study participants resided in adjacent regions (intervention region included Minneapolis, MN, USA and the usual care region included St Paul, MN, USA and adjacent areas in MN, USA) separated by natural geographic barriers. Eligible patients were identified through a review of the Medica plan’s administrative claims database and were asked to complete the Migraine Therapy Assessment Questionnaire (MTAQ), a tool to assess the presence of migraine management issues. Responders (patients completing the MTAQ) who reported ≥2 migraine care risk indicators in the intervention region received the disease management intervention, while all responders in the control region received usual care. Responders in both regions with ≥2 migraine care risk indicators were also asked to complete the Migraine Disability Assessment (MIDAS) questionnaire and the 12-item Short Form Health Survey (SF-12) at baseline. All participants were asked to complete the same instruments at the end of the program as they did at baseline. Interventions: Disease management intervention activities included patient and physician education, as well as a patient feedback report containing migraine care risk indicators, which was sent to the treating physicians. All patients were encouraged to contact their physicians if they reported two or more migraine management issues. Results: A total of 2232 patients with migraine were enrolled in the study (1373 from the intervention region and 859 in the control region). Compared with patients in the control region, patients in the intervention region reported, as measured by MTAQ, significant improvement in migraine symptom relief, more knowledge about potential migraine triggers, a decrease in economic burden, and more satisfaction with migraine treatment. In addition, the change in MIDAS scores from baseline showed a greater shift towards decreasing disability in the intervention group compared with the control group. However, no statistically significant improvement was detected in terms of health status as measured by SF-12. Of the participants in the intervention region who completed the program evaluation survey, 40% indicated that they called or visited their physicians regarding their migraine survey results if it was recommended. For those contacting their physicians, 76% had their medications changed and 75% noted an improvement in relief due to a change in medication. Conclusions: Considering the significant toll of migraine on patients, employers, and the healthcare system, healthcare plans should consider implementing migraine disease management programs to improve migraine care.     

Outcomes Associated with a Disease Management Program for End-Stage Renal Disease

Disease-state management is gaining in use for the management of chronically ill individuals including those with diabetes mellitus, congestive heart failure, asthma, and some forms of cancer. Recently, disease management (DM) has been applied to patients with chronic kidney disease (CKD), a growing population of patients with high annual costs. CKD is ideally suited to DM since the definition of the condition is unambiguous and current care is highly fragmented. There are currently over 240 000 patients receiving dialysis for end-stage renal disease (ESRD), with projected numbers of nearly 600 000 by 2010, and nearly 9 million individuals with CKD not yet on dialysis. The total cost of care for patients with ESRD alone exceeded $US17 billion in 2000. Over 40% of costs for patients with ESRD result from hospitalizations, many of which can be avoided. In addition, much of the clinical morbidity and cost relates to associated comorbidities rather than ESRD per se, with little management presently provided for these conditions in the dialysis facility setting.DM for CKD uses field-based nurse care managers who can risk-assess patients and provide coordination of care so that the renal issues as well as comorbidities are identified and appropriately managed. Although few results from such efforts have been published, those that have, from RMS Disease Mnagement Inc., show remarkable improvements in a variety of clinical outcomes including mortality and hospitalization.Challenges to expanding DM for CKD include up-front funding to provide the needed DM, the availability of robust information systems to manage and analyze clinical and financial data, and the interest and participation of nephrologists, primary care providers and dialysis facilities, as well as other key providers to ensure that the DM approach is effective. With continuing increases in the number of patients with CKD in managed health plans, DM for this population will be even more important in the future to optimize clinical outcomes while constraining the costs of care.     

Cost Analysis of a Maternity Disease Management Program

Background

With over 4 million births annually in the US, pregnancy is a high-volume and high-cost condition for both public and private payers. Although pregnancy is generally considered to be a time of Wellness, certain maternal habits and medical conditions place a pregnancy at risk of maternal and/or neonatal complications. Disease management (DM) provides a framework in which to identify and manage pregnancies at risk of expensive adverse outcomes.

Objective

To analyze the costs of a maternity DM program in a commercial health plan population.

Methods

We conducted an analysis of maternal and newborn data from a commercial health plan collected before utilization of a maternity DM program (baseline period: January–December 2003) and after initiation of the program (operations period: January–December 2004). The maternity DM program consisted of telephonic risk assessment, patient education, and case management of patients identified as being at high risk for adverse pregnancy outcomes. The average costs per high-risk case and high-risk condition were compared between the baseline and operational periods.

Results

A baseline population of 8704 pregnant women in a delivery window of 1 January 2003–31 December 2003 was compared with 7770 pregnant women enrolled for maternity DM in the operations window of 1 January 2004–31 December 2004. In the baseline period, 4756 high-risk conditions were found in 3626 of the 8704 women who experienced a pregnancy (41.7% of patients; a mean of 1.3 conditions per affected pregnancy), while in the operational period, 4377 high-risk conditions were found in 3352 of the 7770 women who experienced a pregnancy (43.1 % of patients; a mean of 1.3 conditions). The most common high-risk condition in both study windows was preterm labor (22.8% baseline vs 23.6% operational). The largest average cost per high-risk case in both periods was for multiple-gestation pregnancies ($US23 171 vs $US28 148; year 2004 values). The cost per mother-baby pair was $US9686 in the baseline period compared with $US9116 in the operations period ($US478 per net savings per mother-baby pair). The net return on investment for maternity DM services was 5.2:1 (DM costs to payer of $US713 220).

Conclusion

This analysis demonstrates that maternity DM in a commercial population is cost saving.

     

“全项目管理”在疾病控制事业管理中的应用

作者通过试行机构层"全项目管理"机制的实践,阐述了通过全部业务的项目化管理以全面落实疾病控制绩效考核,提高工作效益的理念。国家通过公共卫生服务内容项目化确保公共卫生均等化目标最终实现的方略,要求疾病控制事业项目化管理的体制机制设计必然从项目责任上形成疾病控制项目筹资方、管理方和执行方立体格局。同时,在确保各项业务项目得到规范实施、科学评价的同时,也能确保各项管理活动得到规范实施和科学评价。也就是说,将项目管理理念运用到疾控机构日常管理的各重要环节中,通过管理的项目化,来推进和实现业务的项目化管理既是现实需要,也是大势所趋。     

An Empirical Model for the Development of Disease Management Programs in General Practice

Objective: The aim of this research was to develop an empirically derived methodological model for the development of disease management programs in general practice. The model was developed applying the example of hypertension. Methods: Using a rapid appraisal framework, empirical investigations, which included structured panel discussions, face-to-face structured interviews and formal presentations, were conducted to design, test and confirm the disease management model. A broad range of stakeholders were consulted from general practice, other health providers, industry, federal government and consumers. Results: The outcome of the research was a testable definition of disease management for hypertension, and the Australian Disease Management Approach to Hypertension in General Practice (ADAGE) for patients at low to high risk of major cardiovascular events. The central platform of this program is an information, communication technology package for risk assessment and management, plus access to a dietician commissioned by the program and a tailored suite of audiovisual and written material. Conclusion: This study used hypertension to demonstrate a model for the development of theoretically sound disease management models for application in general practice. The ADAGE program is currently under evaluation in a cluster randomized trial of 76 general practices in South Australia and Victoria in both metropolitan and rural settings. It is anticipated that this model will contribute to the development of other disease management programs for addressing chronic disease.     

One-Year Outcomes from a Disease Management Program for Chronic Obstructive Pulmonary Disease

Objective: to present the results of our 1-year, telephonic disease management interventions for 349 patients with mild to severe chronic obstructive pulmonary disease (COPD). Methods: parameters measuring utilization of medical services, days lost from work, and quality-of-life measurements (utilizing the St. George’s Questionnaire) were determined for the 12-month period prior to enrollment of patients and compared with those observed during the 12-month participation in the program. Results: there was a statistically significant reduction in all utilization measurements and a statistically significant improvement in quality of life. Emergency room (ER) visits decreased by 57% (p < 0.001), hospitalizations by 53% (p < 0.001), intensive care unit admissions by 66% (p = 0.001), unscheduled office visits by 67% (p < 0.001), and oral antibacterial bursts by 48% (p < 0.001). Of the 114 employed patients, days missed from work were reduced by 77% (p < 0.001). The total saving from reduction in hospitalizations and ER visits was $US672 000. This was against an approximate cost of the program of $US223 500 (average of $US635 per enrollee). Costs associated with medications and physician visits were not obtained. Conclusions: these outcomes suggest that, with ongoing patient support, the provision of physical rehabilitation, and improved communication between those engaged in the healing process, it is possible to reduce utilization and overall healthcare expenditures and improve the quality of life for a population of patients with moderate to severe COPD. This intervention was both cost effective and medically effective. Our experience suggests that further trials of COPD disease management are warranted.     

Improved Diabetes Control Through a Provider-Based Disease Management Program

Objective: The primary aim of this study was to evaluate the effectiveness of a newly implemented hospital-based diabetes mellitus disease management program. A secondary aim was to determine if relationships existed among variables.Design and setting: Effectiveness was evaluated in terms of glycemic control, post-program acute care resource utilization, adherence with American Diabetes Association (ADA) standards of care, and health-related quality of life. Participants in the Diabetes LifeCare program (DLC) received all standards of care which included diabetes self-management education, medical management by a primary care provider (PCP) supported by an evaluation and recommendations by an Advanced Practice Registered Nurse (APRN), nutritional counseling and at minimum, quarterly follow-up appointments for 1 year.Patients: Patients who were aged ≥18 years and referred to the DLC program with new a diagnosis or history of type 1 or 2 diabetes mellitus.Main outcome measures and results: Results demonstrated that at 3 months after enrollment in the DLC, participants in the study with available data (n = 142) had a mean decrease in predominant glycosylated hemoglobin (HbA_1c) values from 9.31 to 7.21 (p < 0.001). The HbA_1c value for participants with data at the 6-month visit (n = 66) decreased from 9.23 to 7.22 at 3 months and to 6.80 at 6 months (p < 0.001). At baseline, 52 of the 142 participants (36.5%) had HbA1c values less than 8.0, compared with 107 patients (75.4%) at 3 months (p < 0.0001). A total of 185 of 227 patients (81.5%) received eye examinations as per ADA guidelines. A total of 225 of 227 patients (99.1%) were in compliance with the ADA guidelines for nutritional counseling as a result of participation in the program. On the Physical Component Summary, their mean scores were 42.75 [standard deviation (SD) = 11.17] at enrollment, compared with 45.12 (SD = 10.52) at 3 months (p < 0.001). The Mental Component Summary score increased from 47.52 (SD = 11.90) to 50.83 (SD = 10.47) [p < 0.001]. Regarding resource utilization, during the follow-up period only 3 of 227 patients had emergency room visits and there were no inpatient visits for acute problems related to diabetes.Conclusions: Our results show that, in the short term, significant improvement in glycemic control can be achieved through a comprehensive program of patient education and management, that includes collaborative efforts with the patient’s primary care provider. High rates of eye and foot examination can also be attained by reinforcing patient involvement and admissions for acute metabolic complications can be minimized.     

An Inner-City Asthma Disease Management Initiative

Objectives: The purpose of this study was to evaluate the effects of a recently implemented disease management programme [Asthma Control and Education Program (ACE)] on patient outcomes (clinical and functional) and on resource utilisation in socio-economically disadvantaged (and largely Hispanic) individuals with asthma treated at Hartford Hospital. Design and Setting: Using standardised measures (i.e. the Health Status Questionnaire [HSQ] and the Center for Epidemiologic Studies Depression Scale) and conducting serial assessments the investigators determined: (i) functional status at intake and follow-up; (ii) change in the level of illness severity over time; (iii) patientss’ acquisition of self-management skills; and (iv) the type and frequency of acute care services utilised pre- and post-enrolment in ACE. Patient Population:The study participants consisted of consenting individuals, aged 18 years or older, enrolled in ACE from 1 January 1997 to 30 September 1998. Each individual had a comprehensive intake interview (data collected included clinical and financial status) followed by 3 educational sessions. Results: While only 34.7% (n = 282) of the 813 patients referred from Hartford Hospital to physicians or emergency department services elected to participate in the programme, 60% of participants completed the educational sessions. 73% of these enrollees returned for the 3-month follow-up.On the HSQ, the mean Physical Composite Summary (PCS) scores increased from 33.3 ± 10.01 at baseline to 41.6 ± 11.48 at the 3-month follow up and 45.3 ± 10.30 at the 6-month follow-up. In an analysis using only those patients (n = 50) with HSQ scores at baseline and 3- and 6-month follow-ups, there were statistically significant increases in both the PCS and the Mental Composite Summary scores (p < 0.001).For the 68 patients with severity data at baseline, 3 months and 6 months there was a statistically significant improvement over time (p < 0.001). For example, 4.4% had severe disease at 6 months versus 69.1% at baseline. A subset of 73 patients for whom pre-ACE data were available realised a 37% reduction in emergency department use and a 52% reduction in in-patient visits. Conclusions: The evaluation of disease management programmes requires outcomes data. The results of this study of an asthma disease management programme indicate there was an improvement in overall functioning, illness severity, self-management, and utilisation of inpatient and emergency department services. Although a cause and effect relationship could not be assumed, the results suggest that the disease management model is an effective one for the studied population of inner-city patients, socioeconomically disadvantaged individuals previously identified as high utilisers of healthcare services and as having significant environmental exposures problematic for patients with asthma. As currently designed, however, this programme does not address the healthcare needs of the large number of referred patients who choose not to enrol or the enrollees (40%) who do not complete the education and follow-up sessions.     

慢性病监测系统后台管理程序的设计与实现

目的:设计实现慢性病监测系统中的后台管理子程序。方法:以Winform平台为基础,按照Web Service架构,将程序分为5个主要功能模块,采用C#语言编程实现。结果:所编写程序能够实现设计功能,人机界面友好,具有良好的扩展性。结论:程序简洁、易用,为慢性病监测系统的设计提供了良好的基础。     

Evaluation of an Asthma Disease Management Program in a Medicaid Population

In 2002, an asthma disease management program was initiated in Washington State in the US. The program was designed for clients of the state’s Medicaid program, which provides health coverage for qualified low-income state residents. In response to the escalating cost of healthcare and because of concerns about the quality of care, the Washington State Legislature mandated implementation of this disease management program as a pilot project to assist individuals to improve their health. Medicaid administrators used a carefully designed process to identify client needs and to obtain proposals for disease management programs.The asthma program seeks to narrow the gap between the standards of care and its practice. In particular, the program aims to provide patients with a richer understanding of asthma and how to control it. This is accomplished through disease education, symptom awareness and management, trigger avoidance, self-monitoring, and education on recommended medication strategies. The program is based on the US National Institutes of Health’s published guidelines on the optimal treatment of asthma.Enrollment of Medicaid clients into the asthma program began in April 2002. This article describes three approaches to evaluation of the first 3 years of the program: (i) 3 years of self-reported client data; (ii) an independent evaluation of the first year’s changes in utilization and quality of care; and (iii) an actuarial analysis of cost effectiveness. The first study used vendor-reported data collected during initial and follow-up assessments. The authors of this first study also reported the results of a satisfaction survey conducted on behalf of the vendor. The independent evaluation conducted by the University of Washington relied on medical record review and claims analysis, and reported statistical analysis of pre/post comparisons. The actuarial study also reported pre/post comparisons using an analysis of claims per member per month in periods before the program started and at 1 and 2 years of program operations.Clients were assessed according to several dimensions of health including self-management, symptoms, functional status, medication management, and trigger management. Numerous interventions were provided to study participants including access to round-the-clock telephone consultation with a registered nurse, self-care education, alerts sent to the primary provider, and symptom issue follow-up. The asthma disease management program outcomes provide evidence of initial success for those clients who completed the reassessment process. The results of the first 3 years of participation in the program indicate trends toward improved health status and client satisfaction with the program. Long-term evaluation will be necessary to determine if the program reduces costs and closes the quality chasm. If successful, this program could serve as a model for programs with similar clients and similar challenges.     

Disease Management and Its Implications for Outpatient Physician Practice

While disease management may be viewed with skepticism, suspicion, or outright hostility, there are several reasons why this approach to care may warrant physician support. The care of patients with chronic illness in usual primary care practice is beset by significant quality gaps, and significant improvement due to other initiatives such as the electronic health record, pay for performance, or consumer-directed healthcare insurance has yet to be consistently demonstrated. In contrast, an emerging body of peer-reviewed literature is demonstrating that disease management does lead to improvements in clinical quality.Disease management is also an intervention that is closely linked to primary care and, thanks to its willingness to assume many of the tasks of patient engagement, can offload and improve practice efficiency. If successful, it can lead to greater reimbursement from pay for performance, and future iterations of disease management may also reimburse physicians for work effort and quality achievements that support its programs. The electronic health record may introduce new challenges, such as a greater reliance on remote messaging, that can also be addressed by this approach to care. It is also possible that future versions of consumer-directed plans will accommodate disease management approaches in an attempt to protect the healthcare consumer with chronic disease from high out-of-pocket expenses. Last but not least, as disease management and the chronic care model evolve, physicians and the sponsors and suppliers of disease management services may find approaches to care that adopt the best elements of each.     

Impact of a Diabetes Disease Management Program by Race and Ethnicity

Background

In the US, diabetes mellitus affects people in all racial and ethnic groups, but the prevalence and risk of complications are considerably higher among African Americans, Hispanics, American Indians, and Alaskan Natives. This study aimed to evaluate the impact of enrollment for at least 1 calendar year in a diabetes disease management program (DDMP) in a large, commercially insured, managed care population. We assessed changes in utilization of preventive services and adoption of diabetes self-management behaviors by race and ethnicity.

Methods

Participants were aged >17 years and had type 1 or 2 diabetes. They were enrolled in a targeted, high-risk level DDMP between January 2003 and September 2003 and were enrolled in the managed care organization (MCO) for a 2-year period beginning 1 year prior to their enrollment in the DDMP. At baseline, 19 483 MCO enrollees who were participating in the targeted high-risk level DDMP were mailed a 40-item, self-administered baseline survey, which took between 10 and 15 minutes to complete. Baseline results are reported elsewhere. One year later, in June 2004, 5174 of the baseline responders were mailed a slightly modified version of the 40-item survey. The survey measured use of eight preventive services (cholesterol test, dental examination, dilated eye examination, urinalysis, foot examination, influenza vaccination, pneumococcal vaccination, and glycated hemoglobin testing) and engagement in four self-management behaviors (blood glucose tests, diet monitoring, exercise, and smoking avoidance).

Results

Of the 5174 follow-up surveys mailed, 1961 (37.9%) were eligible for comparative analysis. Blacks and Hispanics reported more annual healthcare visits (average of 6.2 and 6.5, respectively) compared with Whites (average of 5.0, p < 0.0001). However, at follow-up, both Blacks and Hispanics had lower utilization rates than Whites for six of the eight preventive services that were measured. At follow-up, both Blacks and Whites were more likely than at baseline to report up-to-date status of influenza and pneumococcal vaccination (p ≤ 0.0001). At follow-up, the racial/ethnic gap in self-management behaviors that was observed at baseline had reduced and in fact had reversed direction for glucose tests, as Blacks were more likely than Whites to routinely test their blood glucose.

Conclusions

These data indicate that DDMP enrollment for at least 1 calendar year had a mixed impact on overall diabetes behaviors and on racial/ethnic disparities in preventive services utilization and self-management behaviors. Further studies are needed to give a clearer understanding of why some diabetic MCO enrollees are less likely to use preventive services, and why disparities remain even in settings where healthcare services are universally available.

     

An HMO-Sponsored Primary Care-Based Disease Management and Case Management Initiative

Objective: To describe the cost savings achieved in a health maintenance organization (HMO)-sponsored primary care-based case management and disease management programs.Methods: The HMO-sponsored programs recruited patients in the primary care setting and relied on clinical guidelines and HMO-employed patient education nurses and case management nurses. Total per member per month (PMPM) charges for medical services and changes in selected clinical outcomes before and after entry into HMO-sponsored case management and disease management programs for actively enrolled participants were compared during the fiscal year January 1, 1998 to November 31, 2000.Interventions: The disease management programs addressed asthma, diabetes mellitus and congestive heart failure (CHF). These programs were based on a network of primary care-based nurse educators and case managers promoting clinical guidelines, appropriate use of the insurance benefit, community-based resources, and communication among all healthcare providers.Setting and participants: This initiative was based in 55 primary care sites serving 295 000 insureds across northeastern and central Pennsylvania, USA.Results: 396 patients with asthma had mean baseline PMPM charges of $US298, which decreased to $US276 PMPM after entry. In those with diabetes mellitus, 3556 patients had a mean baseline PMPM charge of $US367 that decreased to $US346. The mean baseline PMPM charge decreased from $US1877 to $US1541 for 1795 patients with CHF. For 3346 patients undergoing case management, the mean baseline PMPM charge was $US1991 and it decreased to $US1545. Total mean reductions in claims over one year of follow-up from the day of entry for patients with asthma, diabetes mellitus, CHF and for case management programs were $US105 544, $US896 112, $US7 237 440 and $US17 907 992, respectively.Additional data regarding asthma-only claims and pre-post days of work loss in the previous six months, diabetes mellitus-only claims and pre-post mean glycosylated hemoglobin A_1c values, inpatient and outpatient CHF total claims, ACE inhibitor use in CHF, and inpatient and outpatient total claims in case management are also provided.Conclusion: While these claims data may be limited by a lack of statistical significance and by regression to the mean, they suggest that case management and disease management programs in asthma, diabetes and congestive heart failure can be associated with significant financial savings compared with baseline levels of utilization. Clinical outcomes data also suggest this approach may be of benefit.     

典型地区慢性病管理的主要做法及成效*

在慢性非传染性疾病管理方面,我国一些典型地区在组织管理、服务流程、服务方式、信息互联互通等方面均进行了积极探索,并取得了较好成效。其中,高血压、糖尿病管理人群血压或血糖控制率较高,规范管理人群控制效果明显高于总体人群。调研发现的问题是,部分地区慢性非传染性疾病管理人群就医费用水平较低,甚至有下降趋势。