Table 1
Characteristics of sampled patients in three collaboration models.
| Characteristic | Direct Management (Wuhan)% (N) | Medical Consortium (Zhenjiang)% (N) | Loose Collaboration (Nanjing)% (N) | Total% (N) | χ2 test |
|---|---|---|---|---|---|
| Sex | |||||
| Male | 42.3% (160) | 44.1% (158) | 49.6% (176) | 45.3% (494) | χ2 = 4.165 |
| Female* | 57.7% (218) | 55.9% (200) | 50.4% (179) | 54.7% (597) | P = 0.125 |
| Age | |||||
| 18–60* | 15.1% (57) | 29.9% (107) | 33.8% (120) | 26.0% (284) | χ2 = 40.923 |
| 61–70 | 44.2% (167) | 41.3% (148) | 34.1% (121) | 40.0% (436) | P = 0.000 |
| 71–80 | 40.7% (154) | 28.8% (103) | 32.1% (114) | 34% (371) | |
| Marital status | |||||
| Married | 91.3% (345) | 93.3% (334) | 93.5% (332) | 92.7% (1011) | χ2 = 1.676 |
| Others* | 8.7% (33) | 6.7% (24) | 6.5% (23) | 7.3% (80) | P = 0.433 |
| Educational level | |||||
| Illiterate/elementary* | 31.0% (117) | 50.8% (182) | 15.5% (55) | 32.4% (354) | χ2 = 137.963 |
| Middle/high school | 52.9% (200) | 40.5% (145) | 50.4% (179) | 48.0% (524) | P = 0.000 |
| College or above | 16.1% (61) | 8.7% (31) | 34.1% (121) | 19.5% (213) | |
| Severity of illness | |||||
| Low | 54.2% (205) | 46.6% (167) | 39.2% (139) | 46.8% (511) | χ2 = 20.926 |
| Medium | 42.3% (160) | 46.1% (165) | 52.4% (186) | 46.8% (511) | P = 0.000 |
| High* | 3.4% (13) | 7.3% (26) | 8.5% (30) | 6.3% (69) | |
| Types of disease | |||||
| Type 2 diabetes mellitus | 19.8% (75) | 24.6% (88) | 31.5% (112) | 25.2% (275) | χ2 = 13.421 |
| Hypertension* | 80.2% (303) | 75.4% (270) | 68.5% (243) | 74.8% (816) | P = 0.001 |
[i] Note: Figures in brackets refer to corresponding frequency.
*Indicates the omitted group in the ordinal logistic regression.
Table 2
Characteristics of sampled health professionals in three collaboration models.
| Characteristic | Direct Management (Wuhan)% (N) | Medical Consortium (Zhenjiang)% (N) | Loose Collaboration (Nanjing)% (N) | Total % (N) | χ2 test |
|---|---|---|---|---|---|
| Sex | |||||
| Male | 34.5% (20) | 20.6% (13) | 22.6% (12) | 26.5% (45) | χ2 = 3.433 |
| Female | 65.5% (38) | 79.4% (50) | 77.4% (41) | 73.5% (129) | P = 0.180 |
| Age | |||||
| 20~30 | 39.7% (23) | 52.4% (33) | 30.2% (16) | 43.6% (72) | χ2 = 7.113 |
| 31~40 | 36.2% (21) | 34.9% (22) | 45.3% (24) | 35.2% (67) | P = 0.130 |
| 41 or over | 24.1% (14) | 12.7% (8) | 24.6% (13) | 21.2% (35) | |
| Education degree | |||||
| College | 30.0% (18) | 47.6% (30) | 39.6% (21) | 40.1% (69) | χ2 = 4.927 |
| Bachelor | 56.9% (33) | 42.9% (27) | 43.4% (23) | 47.1% (83) | P = 0.29 |
| Postgraduate | 12.1% (7) | 9.5% (6) | 17.0% (9) | 12.8% (22) | |
| Occupation | |||||
| Physician | 48.3% (28) | 39.7% (25) | 50.9% (27) | 46.0% (80) | χ2 = 1.655 |
| Nurse | 51.7% (30) | 60.3% (38) | 49.1% (26) | 54.0% (94) | P = 0.437 |
| Professional title | |||||
| Senior | 43.1% (25) | 57.1% (36) | 43.4% (23) | 48.3% (84) | χ2 = 7.946 |
| Middle level | 51.7% (30) | 42.9% (27) | 47.2% (25) | 47.1% (68) | P = 0.080 |
| Junior or below | 5.2% (3) | 0 (0) | 9.4% (5) | 4.6% (8) |
Table 3
Comparisons of structure of three collaboration models.
| Dimensions/indicators | Direct Management Model (Wuhan) | Medical Consortium Model (Zhenjiang) | Loose Collaboration Model (Nanjing) |
|---|---|---|---|
| Ownership of community health centre | Same (the local government) | ||
| Oversight roles over community health centre | Same (the local government) | ||
| Operation responsibilities over community health centre | Hospital | Community health centre | Community health centre |
| Financing source of community health centre | Hospital, Local government, Medical service and other revenue | Local government, Medical service and other revenue | Local government, Medical service and other revenue |
| Personnel recruitment and management over community health centre | Hospital | Community health centre | Community health centre |
| Collaboration implementation cost | +++ | +++ | + |
| Degree of shared financial resources | ++ | ++ | + |
| Degree of healthcare resources integration | +++ | ++ | + |
| Extent of interagency integration | +++ | ++ | + |
| Degree of patient information sharing between hospital and community health centres | ++ | ++ | + |
| Degree of sharing of organisational culture | +++ | ++ | + |
| Adherence to public health service goal | + | ++ | +++ |
| Organisational structure: Administrative authority over community health centres | Hospital | Hospital | community health centres |
| Financing | |||
| Average annual subsidy for hospital from local government from 2009 to 2011 (Yuan) | 1.1 Million | 0 | 0 |
| Facility/equipment: | |||
| 1 Average number of beds at each community health centres | 79.5 | 28.7 | 79.3 |
| 2 Average number of beds at hospital | 800 | 1030 | 1000 |
| 3 Establishing an information-sharing platform between hospital and community health centres | Yes | Yes | Yes |
| Staffing: | |||
| Special office and staff to oversee collaboration and management | Yes | Yes | No |
| Policy | |||
| 1 Incentives to promote institutional collaboration | Yes | Yes | Yes |
| 2 Whether to have health insurance incentives | No | Yes | No |
[i] Note: ‘+‘means least or weakest; ‘+++’ means most or strongest.
Table 4
Indicators of the process of collaboration in the three models, 2011.
| Dimension | Indicator | Direct Management % (N) | Medical Consortium % (N) | Loose Collaboration % (N) | χ2 test |
|---|---|---|---|---|---|
| Medical care | Average number of hospital departments participating in collaboration with community health centres | 21 | 7 | 8 | – |
| Average number of outreach specialists from hospital providing medical care to patients at community health centres | 4 | 3.7 | 3.3 | – | |
| Average amount of time (days) spent by each outreach specialist at community health centres per week | 1.9 | 1.8 | 1 | – | |
| Workload | Percentage increase in community health centre outpatients between 2008 and 2011 (%) | 45.2% | 79.9% | 40.3% | – |
| Training of health workers | Average number of health professionals per year from community health centres receiving training in hospital in 3 years from 2009 to 2011 | 8 | 3.7 | 2.7 | – |
| Referral | Percentage increase in patient referrals to hospitals from community health centres between 2008 and 2011 | 86.7% | –* | 26.9% | – |
| Percentage increase in patient referrals to community health centres from hospitals between 2008 and 2011 | 133.3% | –* | 47.4% | – | |
| Stakeholders’ perceptions | Percentage of health professionals who were aware of the nature of the collaboration | 91.4% (53) | 60.3% (38) | 56.6% (30) | χ2 = 19.777 P = 0.000 |
| Percentage of patients who knew about the nature of hospital–community health centre collaboration. | 20.8% (71) | 24.5% (61) | 41.4% (144) | χ2 = 39.014 P = 0.000 | |
| Percentage of patients who thought community health centre doctors knew their medical history | 45.7% (156) | 69.5% (173) | 61.8% (215) | χ2 = 36.522 P = 0.000 | |
| Percentage of patients who thought specialists from hospitals knew their medical history | 31.4% (107) | 27.7% (69) | 25.6% (89) | χ2 = 2.911 P = 0.233 |
[i] Note: Figures in brackets are corresponding frequencies.
*Data are not available due to lack of record.
Table 5
Patients’ and health professionals’ assessments on health and other outcomes.
| Patients’ perception on communications between hospitals and related community health centres regarding patients’ illness | ||||
|---|---|---|---|---|
| Sufficient % (N) | Not sufficient % (N) | |||
| Direct Management Model | 18.9 (21) | 81.1 (90) | χ2 = 1.081 | |
| Medical Consortium Model | 16.4 (27) | 83.6 (138) | P = 0.583 | |
| Loose Collaboration Model | 14.3 (25) | 85.7 (150) | ||
| Sub-Total | 16.2 (73) | 83.8 (378) | ||
| Patients’ satisfaction with treatment | ||||
| Satisfied % (N) | Medium % (N) | Dissatisfied % (N) | χ2 test | |
| Direct Management Model | 57.4 (220) | 37.1 (142) | 5.5 (21) | χ2 = 4.111 |
| Medical Consortium Model | 64.1 (236) | 31.3 (115) | 4.6 (17) | P = 0.391 |
| Loose Collaboration Model | 61.6 (229) | 32.5 (121) | 5.9 (22) | |
| Sub-Total | 61.0 (685) | 33.7 (378) | 5.3 (60) | |
| Health professionals’ satisfaction with collaborative treatment of patients,% (N) | ||||
| Direct Management Model | 69.0 (40) | 25.9 (15) | 5.1 (3) | χ 2 = 8.094 |
| Medical Consortium Model | 44.5 (28) | 44.4 (28) | 11.1 (7) | P = 0.083* |
| Loose Collaboration Model | 50.0 (26) | 42.3 (22) | 7.7 (4) | |
| Sub-Total | 54.3 (94) | 37.6 (65) | 8.1 (14) | |
| Health professionals’ satisfaction with referrals of patients,% (N) | ||||
| Direct Management Model | 77.6 (45) | 22.4 (13) | 0.0 (0) | χ 2 = 16.625 |
| Medical Consortium Model | 54.0 (34) | 36.5 (23) | 9.5 (6) | P = 0.001* |
| Loose Collaboration Model | 48.1 (25) | 50.0 (26) | 1.9 (1) | |
| Sub-Total | 60.1 (104) | 35.8 (62) | 4.1 (7) | |
| Health professionals’ satisfaction with communication about patients’ illness,% (N) | ||||
| Direct Management Model | 62.1 (36) | 34.5 (20) | 3.4 (2) | χ 2 = 7.908 |
| Medical Consortium Model | 41.3 (26) | 44.4 (28) | 14.3 (9) | P = 0.91* |
| Loose Collaboration Model | 44.2 (23) | 44.2 (23) | 11.6 (6) | |
| Sub-Total | 49.2 (85) | 41.0 (71) | 9.8 (17) | |
[i] *Results of Fisher’s exact test.
Table 6
Parameter estimates (ordinal logistic regression).
| 95% CI | ||||
|---|---|---|---|---|
| Variables | B (SE) | OR | Lower bound | Upper bound |
| Sex | ||||
| Male | –0.159 (0.130) | 0.853 | 0.662 | 1.100 |
| Age | ||||
| 61–70 | 0.211 (0.167) | 1.235 | 0.890 | 1.714 |
| 71–80 | 0.205 (0.158) | 1.228 | 0.900 | 1.673 |
| Marital status | ||||
| Married | –0.110 (0.237) | 0.896 | 0.563 | 1.424 |
| Education level | ||||
| Middle/high school | 0.105 (0.153) | 1.111 | 0.824 | 1.499 |
| College and above | 0.177 (0.196) | 1.194 | 0.812 | 1.755 |
| Severity of illness | ||||
| Low | 1.667*** (0.253) | 5.296 | 3.224 | 8.707 |
| Medium | 0.983*** (0.247) | 2.672 | 1.649 | 4.334 |
| Types of disease | ||||
| Type 2 diabetes mellitus | –0.073 (0.143) | 0.930 | 0.703 | 1.230 |
| Modela | ||||
| Medical Consortium Model | 0.129 (0.164) | 1.138 | 0.825 | 1.570 |
| Direct Management Model | –0.341** (0.157) | 0.711 | 0.522 | 0.967 |
[i] Note: aLoose Collaboration Model was taken as a controlled group.
**Indicates significance at 5%.
***Indicates significance at 1%.