目的

运用数据包络分析（data envelopment analysis, DEA）分别评价2012-2016年我国规模相当的15家三级妇幼保健院和规模相当的36家三级综合医院的效率特征及变动情况，分析影响效率的因素，为改善样本医院效率提供参考。

方法

分别运用C²R-DEA模型计算样本医院技术效率，运用Super-C²R模型计算技术超效率，运用BC²-DEA模型计算样本医院纯技术效率和规模效率，运用Super-BC²模型计算纯技术超效率，运用Cost-DEA模型计算样本综合医院成本效率和配置效率，通过Malmquist指数模型分析样本医院生产率在2012-2016年的变化趋势。效率分析所用软件为DEAP2.1和DEA-SOLVER（LV3.0）。通过描述性统计分析方法描述样本医院基本情况，通过多元线性回归分析效率影响因素，所用软件为SPSS21.0。

结果

样本妇幼保健院部分

1.样本妇幼保健院在2012-2016年期间，在岗职工数、执业医师数、实有床位数、总诊疗人次、出院人次、每职工诊疗人次、每职工出院人次、病床周转次数、平均住院日、医疗收入、药品收入、总收入、总支出、收支结余的年平均增长率分别为6.06%、11.83%、10.03%、10.01%，10.56%、2.91%、3.87%、-0.72%、-2.78%、16.16%、12.52%、17.01%、15.99%、23.28%。

2.产出指标为出院人次和总诊疗人次，投入指标为执业医师数和实有床位数构建的样本妇幼保健院效率评价模型较为科学合理。

3.2012-2016年样本妇幼保健院的技术效率、纯技术效率、规模效率平均值区间分别为0.85-0.89、0.91-0.95、0.92-0.95，技术效率和规模效率增减趋势保持一致。

4.2012-2016年，33.33%-53.33%的样本妇幼保健院处于规模收益递减阶段，处于规模报酬递增的样本医院比例为6.67%-20.00%。

5.2012-2016年样本妇幼保健院的技术效率变动、技术水平变动、纯技术效率变动规模效率变动年平均增长幅度分别为0.5%、-0.7%、0.5%、-0.1%，全要素生产率年均降低了0.2%。

6.提高病床周转次数可提高样本妇幼保健院技术效率、纯技术效率、规模效率；增加每职工出院人次和每职工诊疗人次可提高样本妇幼保健院技术效率、规模效率；减少患者平均住院费用可以提高样本妇幼保健院规模效率；提高每职工住院床日可以提高样本妇幼保健院纯技术效率。

样本综合医院部分

7.2012-2016年样本综合医院的在岗职工数、执业医师数、实有床位数、总诊疗人次、出院人次、每职工诊疗人次、每职工出院人次、病床周转次数、平均住院日、医疗收入、药品收入、总收入、总支出、收支结余年平均增长率分别为5.27%、6.01%、5.04%、7.56%、9.48%、0.92%、3.36%、7.09%、-3.61%、18.52%、10.25%、12.98%、13.39%、3.24%。

8.产出指标为出院人次和总诊疗人次，投入指标为在岗职工数和实有床位数构建的样本综合医院效率评价模型较为科学合理。

9.2012-2016年样本综合医院技术效率、纯技术效率、规模效率平均值区间分别为0.78-0.82、0.84-0.87、0.90-0.95，样本综合医院的纯技术效率小于规模效率。

10.2012-2016年样本综合医院处于规模报酬递减和处于规模报酬递增阶段的比例分别41.67%-77.78%、8.33%-50.00%，绝大部分样本医院处于规模效益无效状态。

11.2012-2016年样本综合医院的技术效率变动、技术水平变动年平均增长幅度分别为年-0.9%、3.9%，全要素生产率年均增加了3.0%。

12.提高样本综合医院病床周转次数可提高样本医院的成本效率、配置效率；增加样本综合医院每职工出院人次可提高样本医院的成本效率、配置效率、纯技术效率、技术效率；提高样本综合医院每职工诊疗人次可提高样本医院的成本效率、纯技术效率、技术效率；提高样本综合医院药品收入占总收入比重可提高样本医院的技术效率；减少样本综合医院患者平均住院日可提高样本医院的成本效率、配置效率、纯技术效率与技术效率；减少样本综合医院患者平均住院费用可以提高样本医院的规模效率、技术效率；提高样本综合医院每职工住院床日可以提高样本医院的配置效率、纯技术效率。

结论

1.2012-2016年样本妇幼保健院与样本综合医院规模呈逐年扩张趋势。2012-2016年大部分样本妇幼保健院和样本综合医院处于规模不适宜状态，由于医院规模过大导致规模效率无效的样本医院比例较大。

2.样本妇幼保健院的技术效率与规模效率变动趋势一致，技术效率受医院规模大小的影响程度较大；样本综合医院纯技术效率明显低于规模效率，说明样本综合医院在提高自身管理水平、发挥技术水平作用与优化组织架构方面仍存在较大改进空间。样本综合医院大幅调整规模后会影响到样本综合医院资产配置比例变化进而影响到配置效率的变化。

3.2012-2016年样本妇幼保健院的全要素生产率均值几乎无变化，技术水平呈下降趋势；2012-2016年样本综合医院的全要素生产率年均增长率为3%，技术进步对生产率改善有明显促进作用。

Objective

Data envelopment analysis (DEA) was used to evaluate the efficiency characteristics, productivity changes and the factors affecting the efficiency of 15 tertiary maternal and child health hospitals with the same scale size and 36 tertiary general hospitals in China with the same scale size in 2012-2016, to provide some scientific basis for the efficiency improvement of sample hospitals.

Methods

Use C²R-DEA model, Super-C²R model, BC²-DEA model, Super-BC² model and Cost-DEA model to calculate the technical efficiency, super technical efficiency, pure technical efficiency, super pure technical efficiency, scale efficiency, cost-efficiency and allocative efficiency of sample hospitals, and use Malmquist productivity index model to calculate the productivity and efficiency changes of the sample hospitals during the period of 2012-2016. Use the software DEAP2.1 and DEA SOLVER(LV3.0) to calucate the DEA efficiency. Statistical analysis mainly include statistical description and multiple regression analysis, use software SPSS17.0 to calucate these statistical analysis.

Results

Part one: sample maternal and child health hospitals

1. During 2012 to 2016, there were an average annual increase of 6.06% in employees of sample maternal and child health hospitals, an average annual increase of 11.83% in practicing physicians, an average annual increase of 10.03% in beds, an average annual increase of 10.01% in outpatient visits, an average annual increase of 10.56% in discharged patients, an average annual increase of 2.91% in outpatient visits per worker, an average annual increase of 3.87% in discharge per worker, an average annual increase of -0.72% in hospital bed turnover, an average annual increase of -2.78% in average day of hospitalization, an average annual increase of 16.16% in medical income, an average annual increase of 12.52% in drug income, an average annual increase of 17.01% in gross income, an average annual increase of 15.99% in gross charge, an average annual increase of 23.28% in balance of revenue and expenditure.

2. The evaluation model of sample maternal and child health hospitals was set by the number of practitioner, the number of beds, the number of discharges and the number of outpatient visits. The model was scientific and reasonable.

3. The average technical efficiency of 15 sample maternal and child health hospitals ranged from 0.85 to 0.89 during 2012 to 2016.In the same period, the average of pure technical efficiency for the sample hospitals ranged from 0.91 to 0.95, the average of scale efficiency for the sample hospitals ranged from 0.92 to 0.95, and the trend of the scale efficiency and the technical efficiency was consistent completely.

4. From 2012-2016, 33.33%-53.33% of the sample maternal and child health hospitals were in the stage of the decreasing returns to scale, and 6.67%-20.00% of the sample hospitals were in the stage of the increasing returns to scale.

5. The Malmquist productivity index model results showed that the technical efficiency of the sample maternal and child health hospitals has increased by 0.5% in annual average during the period of 2012 to 2016, the technical level has been reduced by 0.7% annually, the pure technical efficiency has increased by 0.5% on annual average, the scale efficiency has declined by 0.1% on annual average, and the total factor productivity has declined by 0.2% on annual average.

6. To increase the number of bed turnover can improve the TE, PTE and SE of sample maternal and child health hospitals; To increase the number of hospital discharges per employee and the number of visits per employee can improve the TE and SE of sample hospitals; Reduce the average hospital cost of patients can improve the SE; improve the hospital bed day for each employee can improve the PTE of the sample hospitals.

Part two: sample general hospitals

7. During 2012 to 2016, there were an average annual increase of 5.27% in employees of sample general hospitals, an average annual increase of 6.01% in practicing physicians, an average annual increase of 5.04% in beds, an average annual increase of 7.56% in outpatient visits, an average annual increase of 9.48% in discharged patients, an average annual increase of 0.92% in outpatient visits per worker, an average annual increase of 3.36% in discharge per worker, an average annual increase of 7.09% in hospital bed turnover, an average annual increase of -3.61% in average day of hospitalization, an average annual increase of 18.52% in medical income, an average annual increase of 10.25% in drug income, an average annual increase of 12.98% in gross income, an average annual increase of 13.39% in gross charge, an average annual increase of 3.24% in balance of revenue and expenditure.

8. The output indicators are the number of discharges and the number of outpatient visits. And the input indicators are the number of employees and the number of beds. The model is scientific and reasonable.

9. From 2012 to 2016, the average technical efficiency of the sample general hospitals was 0.78 to 0.82. In the same period, the average pure technical efficiency of the sample hospitals was 0.84 to 0.87, the average scale efficiency was 0.90 to 0.95, the pure technical efficiency was less than the scale efficiency.

10. From 2012 to 2016, 41.67%-77.78% of sample general hospitals were in the stage of the decreasing returns to scale, 8.33%-50.00% of sample hospitals were in the stage of the increasing returns to scale, almost of them were in ineffective scale efficiency because of oversize.

11. The Malmquist productivity index model results showed that the technical efficiency of the sample general hospitals in 2012 to 2016 decreased by 0.9% in annual average, the technical level increased by an average of 3.9%, and the total factor productivity increased by 3.0% annually.

12. To increase the turnover of hospital beds can improve the AE and CE of sample hospitals; to increase the number of hospitalizations per employee can improve the TE, PTE, AE and CE of sample hospitals; to increase the number of visits per employee can improve TE, PTE and CE of sample hospitals; to increasing pharmaceutical revenue can improve the TE of sample hospitals; to reduce average patient hospitalization days can improve the TE, PTE, AE and CE of sample hospitals; to reduce the average cost of  patients hospitalization can improve the TE and SE of the sample hospitalS; to increase the hospital bed days per employee can improve the PTE and AE of the sample hospital.

Conclusions

1. The scale of the sample maternal and child health hospitals and the sample general hospitals has expanded year by year in 2012-2016 years. Most of the sample general hospitals and the sample maternal and child health hospitals were in the scale efficiency ineffectiveness status, and almost of them were in ineffective scale efficiency because of oversize.

2. The trend of scale efficiency and technical efficiency was consistent of sample maternal and child health hospitals, the scale was a key influence factor for technical efficiency, the pure technical efficiency of the sample general hospitals was lower than the scale efficiency. Obviously, it showed that the sample general hospitals still has great improvement space in improving its management level, playing the role of technical level and optimizing the organization structure. The scale is an important factor affecting the allocation efficiency of the sample general hospitals, the allocation efficiency has changed significantly when most of the sample hospitals have scale adjustment remarkablely.

3. The total factor productivity of the sample maternal and child health hospitals almost nothing changed since 2012 to 2016. The total factor productivity of sample general hospitals has increased by 3.0%, the technical changed has improved productivity obvious.