Levels, Trends and Determinants of Technical Efficiency of General Hospitals in Uganda: Data Envelopment Analysis and Tobit Regression Analysis

DOI: https://doi.org/10.21203/rs.3.rs-23855/v3

Abstract

Background: General hospitals provide a wide range of primary and secondary healthcare services. They accounted for 38% of government funding to health facilities, 8.8% of outpatient department visits and 28% of admissions in Uganda in the financial year 2016/17. We assessed the levels, trends and determinants of technical efficiency of general hospitals in Uganda from 2012/13 to 2016/17.

Methods: We undertook input-oriented Data Envelopment Analysis to estimate technical efficiency of 78 general hospitals using data abstracted from the Annual Health Sector Performance Reports for 2012/13, 2014/15 and 2016/17.  Trends in technical efficiency was analysed using Excel while determinants of technical efficiency were analysed using Tobit Regression Model in STATA 15.1.

Results: The Average Constant Returns to Scale, Variable Returns to Scale and Scale Efficiency of general hospitals for 2016/17 were 49% (95% CI, 44% - 54%), 69% (95% CI, 65% - 74%) and 70% (95% CI, 65% - 75%) respectively. There was no statistically significant difference in the efficiency scores of public and private hospitals. Technical efficiency generally increased from 2012/13 to 2014/15, and dropped by 2016/17. Some hospitals were persistently efficient while others were inefficient over this period. Hospital size, geographical location, training status and average length of stay were statistically significant determinants of efficiency at 5% level of significance.

Conclusion: The 69% average variable returns to scale technical efficiency indicates that the hospitals could generate the same volume of outputs using 31% (3,439) less staff and 31% (3,539) less beds. Benchmarking performance of the efficient hospitals would help to guide performance improvement in the inefficient ones. There is need to incorporate hospital size, geographical location, training status and average length of stay in the resource allocation formula and adopt annual hospital efficiency assessments. 

Full Text

This preprint is available for download as a PDF.

Tables

Table 1: Number of health facilities in Uganda disaggregated by level of care 

Facility level 

Count 

Clinics

1,578 (22.75%)

Level II Health Center (HCII)

3,364 (48.49%)

Level III Health Center (HCIII)

1,569 (22.62%)

Level IV Health Center (HCIV)

222 (3.2%)

General hospitals

163 (2.35%)

Regional referral hospitals

13 (0.19%)

Referral hospitals

3 (0.04%)

National referral hospitals

2 (0.03%)

Specialized hospitals

23 (0.33%)

Source: Ministry of Health. Facility Master List, 2018

 

Table 2: Study variables and data sources

 

Variable 

Definition 

Measurement

Source (s) of data

Inputs

 

 

 

Bed

Hospital beds

Total number of beds in the year

AHSPR for FYs 2012/13, 2014/15 and 2016/17

Staff

Medical personnel

Total number of staff (Medical Officers, Dental, Pharmacy, Nursing, Allied Health Professionals, Administrative and Other Staff) in the year

AHSPR (FY 2016/17), Integrated Human Resource Information System; Reports of the Catholic and Protestant Medical Bureaux

Outputs

 

 

 

 OPD

OPD visits

Total number of outpatient visits in the year

AHSPR for FYs 2012/13, 2014/15 and 2016/17

 ADM 

Hospital admissions 

Total number of inpatient admissions

AHSPR for FYs 2012/13, 2014/15 and 2016/17

Deliveries

Deliveries (births)

Total number of deliveries in the year

AHSPR for FYs 2012/13, 2014/15 and 2016/17

Predictors

 

 

 

 Ownership

Hospital ownership 

Authority that owns the hospital: public (1) or private (0)

AHSPR (FY 2016/17)

Hospsize

Hospital size 

Size of the hospital classified using the median number of beds: large (>120 beds [1]), small (<=120 beds [0]). Given the variability in sizes of general hospitals across the world, lack of global or national benchmark for their optimal size and the need to ensure fair distribution of small and large hospitals presentation in the 2 groups, the authors used the median bed size rounded to the nearest ten i.e. 120 as the benchmark to classify the 78 general hospitals as small and large.

PR (FY 2016/17)

 Propqualstaff

Proportion of qualified staff 

Number of staff with formal qualifications (Medical Officers, Dental, Pharmacy, Nursing, Allied Health Professionals, Administrative and Other Staff) as a proportion of all staff in the year

iHRIS, Reports from Catholic and Protestant Medical Bureaus

 Region

Geographical location 

Region where the hospital is located: Central or Western Uganda (1), Northern or Eastern Uganda (0)

AHSPR (FY 2016/17)

 BOR

Bed occupancy rate 

Total annual inpatient days as a ratio of annual available bed days ×100

AHSPR (FY 2016/17)

TrainingStatus

Training status 

Hospital is used for training health professionals or not: Yes (1) and No (0) 

Ministry of Health Training Unit, Catholic Medical and Protestant Medical Bureaus

 OPDIBD

Outpatient visit to total inpatient days ratio

Total number of OPD visits divided by total number of inpatient bed days in the year

AHSPR (FY 2016/17)

 AvStayADM

Average length of stay 

Total annual number of inpatient days spent/total annual number of admissions

AHSPR (FY 2016/17)

 

Table 3: Descriptive statistics for input and output variables

Group

Variable

obs

Mean

Std. Dev

Min

Max

A - All hospitals

Beds

78

142

58.75402

61

305

Staff

78

146

67.14357

42

433

OPD

78

38720

27127.02

4873

178146

ADM

78

8,948

4540.011

1427

23560

Deliveries

78

2,127

1438.253

229

7002

B – Public hospitals

Beds

40

126

38.33315

76

224

Staff

40

135

28.97035

81

204

OPD

40

53,562

28703.54

18790

178146

ADM

40

10,972

3882.417

3885

23560

Deliveries

40

2,776

1637.149

544

7002

C – PNFP hospitals

Beds

38

159

70.99094

61

305

Staff

38

159

90.5376

42

433

OPD

38

23,097

13197.76

4873

64580

ADM

38

6,817

4231.765

1427

20446

Deliveries

38

1,444

738.1052

229

3453

 

Table 4: Descriptive statistics for the continuous independent variables

Variable

obs

Mean

Std. Dev

Min

Max

Propqualstaff

78

76.05128

12.38652

43

100

BOR

78

66.61538

36.18503

15

178

OPDIBD

78

1.371795

0.9109575

0.2

4.5

AvStayADM

78

3.871795

1.399063

2

8

 

Table 5: Descriptive statistics for the categorical independent variables

Variable

Coding

Frequency

Percent

Cumulative %

Ownership

1, Public

40

51.28

51.28

0, PNFP

38

48.72

100

Hospital size

1, Big (>120 beds)

37

47.44

47.44

0, Small (<= 120 beds)

41

52.56

100

Geographical location

1, Central or Western

42

53.85

53.85

0, Northern or Eastern

36

46.15

100

Training status

1, Yes

28

35.90

35.90

0, No

50

64.10

100

 

Table 6: Hospital efficiency scores disaggregated by hospital ownership during FY 2016/17

Parameter in separate groups of hospitals

CRS TE 2016/17

VRS TE 2016/17

SE 2016/17

All hospitals

 

 

 

Number of efficient hospitals

2

8

2

Number of inefficient Hospitals

76

70

76

Efficient hospitals (%)

3

11

3

Inefficient hospitals (%)

97

90

97

Average efficiency score (%)

49

69

70

Minimum score (%)

13

25

18

Maximum score (%)

100

100

100

Public hospitals

 

 

 

Number of efficient hospitals

2

6

2

Number of inefficient hospitals

38

34

38

Efficient hospitals (%)

5

15

5

Inefficient hospitals (%)

95

85

95

Average efficiency score (%)

64

82

78

Minimum score (%)

28

50

45

Maximum score (%)

100

100

100

PNFP hospitals

 

 

 

Number of efficient hospitals

10

16

10

Number of inefficient hospitals

28

22

28

Efficient hospitals (%)

26

42

26

Inefficient hospitals (%)

74

58

74

Average efficiency score (%)

73

83

87

Minimum score (%)

28

43

30

Maximum score (%)

100

100

100








 

Table 7: Top 10 hospitals in FYs 2012/13, FY 2014/15 and FY 2016/17

FY 2012/13

FY 2014/15

FY 2016/17

SN

Hospital

SE Score (Super-effCRS)

Hospital

SE Score (Super-effCRS)

Hospital

SE Score (Super-effCRS)

1

Iganga

1.0000 (2.0224)

Iganga

1.0000 (1.800) 

Iganga

1.0000(1.977)

2

Busolwe

1.0000(1.4456)

Busolwe

1.0000 (1.373) 

Tororo

1.0000 (1.112)

3

Bwera

1.0000 (1.2163)

Mityana

1.0000 (1.087) 

Kalongo

0.9956

4

Mityana

1.0000 (1.0771)

Kagadi

1.0000 (1.027) 

Kitgum

0.9946

5

Masafu

1.0000 (1.0768)

Pallisa

1.0000 (1.016) 

Mityana

0.9818

6

Tororo

0.9910

Ibanda

0.9958 

Angal St. 

Luke

0.9795

7

Kitagata

0.9906

Tororo

0.9926 

Bududa

0.9759

8

Moyo

0.9896

Kitgum

0.9888 

Atutur

0.9752

9

Ibanda

0.9884

Angal 

St. Luke

0.9880 

Entebbe

0.9661

10

Entebbe

0.9817

Nebbi

0.9744 

Ibanda

0.9584

 

 

 

Table 8: Bottom 10 hospitals in FYs 2012/13, FY 2014/15 and FY 2016/17

FY 2012/13

FY 2014/15

FY 2016/17

SN

Hospital

CRS TE 

Hospital

CRS TE 

Hospital

CRS TE 

1

Matany

0.19

Aber

0.43

St. Francis Nyenga

0.13

2

Maracha

0.21

Abim

0.24

Amai Community

0.16

3

St. Francis

Nyenga

0.21

Amai Community

0.30

Kiwoko

0.17

4

Kisiizi

0.21

Amudat

0.45

Virika

0.17

5

Buluba - Leprosy

0.23

Anaka

0.46

Kisiizi

0.18

6

Rugarama

0.24

Angal St. Luke

0.54

Rushere Community

0.19

7

St. Anthony's 

Tororo

0.24

Apac

0.64

Buluba - Leprosy

0.19

8

St. Joseph Kitovu

0.25

Atutur

0.96

Villa Maria

0.19

9

Abim

0.25

Bududa

0.62

Nkokonjeru

0.23

10

Virika

0.26

Bugiri

0.64

St. Anthony's Tororo

0.23

 

Table 9: Output of Tobit Regression 

VRSDEAIneffScore

Coef

Std. Error

t

p>(t)

[95% Conf. Interval]

Lower               Upper

Ownership

-.2373533

.1566698

-1.51

0.134

-.5498213

.0751147

Hospsize

.3064594

.1306399

2.35

0.022**

.0459062

.5670126

Propqualstaff

-.0090919

.0050784

-1.79

0.078

-.0192205

.0010367

Geographical location

.2581932

.1243699

2.08

0.042**

.0101453

.5062411

BOR

-.0016612

.0021884

-0.76

0.450

-.0060258

.0027035

TrainingStatus

.2620071

.1297391

2.02

0.047**

.0032505

.5207637

OPDIBD

.0816522

.085905

0.95

0.345

-.0896799

.2529843

AvStayADM

.1948153

.0580116

3.36

0.001**

.0791149

.3105158

_cons

.2592612

.4617965

0.56

0.576

-.6617628

1.180285

Var (e. VRSDEAIneffScore

.2232797

.0383208

 

 

.1585589

.3144181

**Statistically significant at 5% level of significance