Top Banner
Projecting the CostEffectiveness of Universal Access to Modern Contraceptives in Uganda Joseph B. Babigumira, MBChB, MS, PhD Pharmaceutical Outcomes Research and Policy Program School of Pharmacy University of Washington Fifth Annual Research Conference on Population, Reproductive Health, and Economic Development January 19 – 21, 2011 Marseilles, France 1
25

Projecting the Cost-Effectiveness of Universal Access to ...

Mar 23, 2023

Download

Documents

Khang Minh
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Projecting the Cost-Effectiveness of Universal Access to ...

Projecting the Cost‐Effectiveness of Universal Access to  Modern Contraceptives in Uganda

Joseph B. Babigumira, MBChB, MS, PhDPharmaceutical Outcomes Research and Policy Program

School of Pharmacy

University of Washington

Fifth Annual Research Conference on Population, Reproductive 

Health, and Economic Development

January 19 –

21, 2011 

Marseilles, France

1

Page 2: Projecting the Cost-Effectiveness of Universal Access to ...

Acknowledgement

Funding from the William and Flora Hewlett Foundation and Institute of 

International Education (IIE) Dissertation Fellowship in Population, 

Reproductive Health, and Economic Development

2

Page 3: Projecting the Cost-Effectiveness of Universal Access to ...

Summary

1.

Introduction

2.

Methods

3.

Results

4.

Discussion

5.

Conclusion

3

Page 4: Projecting the Cost-Effectiveness of Universal Access to ...

4

Average 6.7

births (annual 

population growth rate: 3.2%)

Only 31%

have access to modern 

effective contraception

Up to 45%

of births in 2006 were 

unplanned 

Have more children (6.7) than 

desired (5.1) 

Unintended pregnancies due to 

lack of contraceptive use (88%) 

and contraceptive failure (12%)

Introduction

Women in Uganda

Page 5: Projecting the Cost-Effectiveness of Universal Access to ...

Contraception may reduce fertility, 

enables family planning, improves 

socioeconomic conditions.

Given potential benefits, policy makers 

should ensure access, but access remains 

poor.

With a total per capita health 

expenditure US$24, Uganda’s 

government‐run healthcare system must 

prioritize.

Is universal access to modern 

contraception a comparatively efficient 

use of scarce resources and should policy 

makers take steps to increase access? 

5

Introduction

Contraception in Uganda

Page 6: Projecting the Cost-Effectiveness of Universal Access to ...

Comparators

The study compared two contraceptive use scenarios: 

1.

The Current Contraceptive Program 

(CCP) in which contraceptive coverage 

would remain at the status quo.

2.

A New Contraceptive Program (NCP)

that 

would provide universal access to modern 

contraception in Uganda.

6

Page 7: Projecting the Cost-Effectiveness of Universal Access to ...

Methods Overview•

A Markov model based on states of sexual activity, contraceptive

use and 

pregnancy was developed to compare the New Contraceptive Program

(NCP) to 

the Current Contraceptive Program (CCP).  

The model followed a hypothetical cohort of 15‐year old girls over a lifetime 

horizon. 

The analysis was performed from both the societal perspective which included all 

costs and governmental perspective which included only the direct medical costs 

incurred by the Ministry of Health. 

Data were obtained from the Uganda National Demographic and Health Survey 

and from published and unpublished sources.  

The main outcomes of the analysis were cost per life‐year (LY) gained, cost per 

disability‐adjusted life‐year (DALY) averted, cost per pregnancy averted, and cost 

per unit of fertility reduction.  

The NCP was considered cost‐effective if the incremental cost‐effectiveness ratio 

(ICER) was less than the 3 times the Uganda’s GDP per capita.

7

Page 8: Projecting the Cost-Effectiveness of Universal Access to ...

Markov Modeling 

Markov modeling is an extension of decision analytic modeling.

Markov models are used to simulate chronic disease processes in which events 

occur multiple times.  In this case, we adapt Markov modeling to

the reproductive 

experience of women in Uganda.

The disease is divided into mutually exclusive health states and

“allowed”

transitions between states are defined.

Transition probabilities are used to estimate the rate of movement between 

health states every cycle.

The model is run for multiple cycles over the time horizon of the analysis, usually 

until all patients end in a “terminal state”

(usually “death”—an “absorbing state”).8

Page 9: Projecting the Cost-Effectiveness of Universal Access to ...

Markov Model

9

Page 10: Projecting the Cost-Effectiveness of Universal Access to ...

Age‐Specific Transition Probabilities (All data from 2006 Uganda Demographic and Health Survey)

Transition 15 – 19 20 – 24 25 – 29  30 – 34  35 – 39  40 – 44  45 – 49 

NSA to INC 0.02  0.17 0.17  0.12  0.08  0.05  0.13 

NSA to UNC 0.07 0.21 0.26  0.29  0.29  0.27  0.19 

NSA/UNC to MOC 0.09 0.17 0.19  0.22  0.19 0.21  0.19 

NSA/UNU to TRC 0.02 0.03  0.04  0.04  0.04  0.06  0.05 

NSA to NSA 0.80 0.41  0.34  0.31  0.39  0.41  0.47 

UNC/INC to PRE 0.85 0.85  0.85  0.83  0.81  0.69  0.17 

UNC to UNC 0.07 0.05  0.05  0.05 0.06  0.16  0.53 

All states to Dead 0.002 0.003  0.006 0.009  0.012  0.011  0.011

NSA

Not Sexually ActiveINC

Intentional Non‐ContraceptionUNC

Unintentional Non‐Contraception

10

MOC

Modern ContraceptionTRC 

Traditional ContraceptionPRE

Pregnant

Page 11: Projecting the Cost-Effectiveness of Universal Access to ...

Other Transition ProbabilitiesProbability Base Case Range Reference

MOC – PRE  0.03 0.02 – 0.03 Trussell (2009)

TRC – PRE  0.20 0.16 – 0.24 Trussell (2009)

MOC – INC  0.25 0.20 – 0.29 Blanc et al. (2009)

MOC – UNC  0.34 0.27 – 0.41 Blanc et al. (2009)

TRC – INC  0.26 0.21 – 0.31 Blanc et al. (2009)

TRC – UNC  0.36 0.27 – 0.41 Blanc et al. (2009)

PRE – NSAϕ 0.73 0.58 – 0.88 2006 UDHS

PRE – INC  0.03 0.02 – 0.04 2006 UDHS

PRE – UNC  0.06 0.05 – 0.08 2006 UDHS

PRE – MOC  0.04 0.03 – 0.05 2006 UDHS

PRE – TRC  0.01 0.01 – 0.02 2006 UDHS

PRE – Deadψ 0.0034 0.0028 –

0.0041 2006 UDHSϕAlso probability of live birth. Calculated by subtracting ectopic pregnancies, induced abortions, miscarriages and still births ψMaternal mortality

11

Page 12: Projecting the Cost-Effectiveness of Universal Access to ...

Other Parameters

Parameter Base Case Range ReferencePregnancy ComplicationsMiscarriage 0.049 0.039 – 0.059 Casterline (1989)Ectopic pregnancy 0.014 0.011 – 0.017 Liskin (1992)Abortion 0.190 0.152 – 0.059 Singh et al. (2005)Still birth 0.017 0.014 – 0.020 Statnton et al. (2006)MortalityNeonatal mortality 0.021 0.017 – 0.025 2006 UDHSInfant mortality 0.055 0.044 – 0.067 2006 UDHSChild mortality 0.049 0.080 – 0.120 2006 UDHSLife expectancy at 2.5 years 51.7 ‐‐ WHO Life TablesDALYs lostMaternal conditions 0.272 0.218 – 0.327 WHO

12

Page 13: Projecting the Cost-Effectiveness of Universal Access to ...

Costs

Cost Base Case($US)

Range($US)

Reference

Contraception (MOH) 25.5 12.7 – 38.2 Weissman et al. (1999); Levin et al. (2003)

Contraception (Societal) 39.0 19.5 – 58.5 Weissman et al. (1999); Levin et al. (2003)

Pregnancy (MOH) 79.4 40.1 –

120.4  Weissman et al. (1999); Levin et al. (2003)

Pregnancy (Societal) 142.2 71.1 –

213.4 Multiple*

Annual productivity loss 354.2 ‐‐ CIA World Fact Book

*Includes  Weissman et al (1999); Levin et al (2003); CIA World Fact Book and Primary Data

13

Page 14: Projecting the Cost-Effectiveness of Universal Access to ...

Results:  Mean Costs and OutcomesCurrent ContraceptiveProgram

New Contraceptive Program

Pregnancies 9.51 7.90Live births 6.92 5.79Life expectancy (Years) 28.65 28.74DALYs 27.01 27.38Ectopic pregnancies 0.13 0.11Miscarriages  0.46 0.39Abortions 1.80 1.51Still births 0.16 0.14Neonatal deaths 0.20 0.17Infant deaths 0.53 0.44Child deaths 0.95 0.79Societal costs ($US) 1,041 1,074Governmental costs ($US) 397 448

14

Page 15: Projecting the Cost-Effectiveness of Universal Access to ...

Results: Cost‐Effectiveness Analysis

Cost ($US) IncrementalCost ($US)

DALYs IncrementalDALYs

ICER ($US/DALY)

CCP 1,041 ‐‐ 27.01 ‐‐ ‐‐

NCP 1,074 33 27.38 0.37 88

1. Societal Perspective

2. Governmental Perspective

Cost ($US) IncrementalCost ($US)

DALYs IncrementalDALYs

ICER ($US/DALY)

CCP 397 ‐‐ 27.01 ‐‐ ‐‐

NCP 448 51 27.38 0.37 138

15ICER – Incremental Cost-Effectiveness Ratio

Page 16: Projecting the Cost-Effectiveness of Universal Access to ...

Results: Other ICERs Comparing NCP vs. CCP

Societalperspective

Governmental perspective

ICER ($/LY) 361 567ICER ($/pregnancy averted) 20 32ICER ($/unit of fertility reduction) 29 45ICER ($/Ectopic pregnancy averted) 1,477 2,323ICER ($/Miscarriage averted) 464 730ICER ($/Abortion averted) 112 176ICER ($/Still birth averted) 1,625 2,555ICER ($/Neonatal death averted) 1,083 1,703ICER ($/Infant death averted) 361 567ICER ($/Child death averted) 203 319

16

Page 17: Projecting the Cost-Effectiveness of Universal Access to ...

Cost‐Consequences Analysis Assuming a Cohort of  100,000 Ugandan Women  

17

CCP NCP DifferenceSocietal costs 104,142,000 107,392,000 3,250,000MOH costs 39,691,000 44,802,000 5,111,000Pregnancies 950,000 790,000 ‐160,000Life years 2,865,000 2,874,000 9,000DALEs 2,701,000 2,738,000 37,000Ectopic pregnancies 13,300 11,100 ‐2,200Induced abortions 180,000 151,000 ‐29,000Miscarriages 46,000 39,000 ‐7,000Still births 16,000 14,000 ‐2,000Live births 692,000 579,000 ‐113,000Neonatal deaths 20,000 17,000 ‐3,000Infant deaths  53,000 44,000 ‐9,000Child deaths 95,000 79,000 ‐16,000

Page 18: Projecting the Cost-Effectiveness of Universal Access to ...

Sensitivity Analysis: Impact on Incremental  Costs

18

Cost of modern contraception ($23‐$90)

Cost of pregnancy ($60‐$240)

Discount rate (0‐5%)

Probability of live birth (0.58‐0.88)

Other less sensitive 

parameters

Incremental Cost (US $)

Page 19: Projecting the Cost-Effectiveness of Universal Access to ...

Sensitivity Analysis: Impact on Incremental DALYs

19

Discount rate (0‐5%)

Probability of discontinuation, modern (27‐41%)

DALYs lost, maternal conditions (0.21‐0.33)

Probability of maternal mortality (0.28‐0.41%)

Probability of continuation, modern (26‐34%)Probability of live birth (58‐88%)

Other less sensitive 

parameters

Incremental Disability‐Adjusted Life Years

Page 20: Projecting the Cost-Effectiveness of Universal Access to ...

Probabilistic Sensitivity Analysis Distribution of Cost‐Effectiveness Pairs on the 

Cost‐Effectiveness Plane

20

Page 21: Projecting the Cost-Effectiveness of Universal Access to ...

Probabilistic Sensitivity Analysis Cost‐Effectiveness Acceptability Curve

21

Page 22: Projecting the Cost-Effectiveness of Universal Access to ...

Discussion

Universal access to modern contraceptives 

would be highly cost‐effective.  

Results confirm findings of earlier studies 

while improving on methods and accounting 

for parameter uncertainty.

Relatively affordable at $70 million annually 

(compared to $450 million for HIV/AIDS 

care).

Would reduce complications, deaths, 

disability and economic impact of  unsafe 

induced abortions.

22

Page 23: Projecting the Cost-Effectiveness of Universal Access to ...

Limitations

A constant contraceptive mix over time was modeled while preferences 

for different methods may change.  Additionally, new contraceptive 

technologies may improve the willingness of women to take up 

contraception.  

The model does not capture changing fertility preferences over time yet 

the desired number of children tends to reduce over time  with increasing 

socioeconomic development and as the percentage  of educated women 

increases.

23

Page 24: Projecting the Cost-Effectiveness of Universal Access to ...

Conclusions

Providing universal access to 

modern contraceptives in Uganda 

appears to represent a 

comparatively efficient use of 

scarce healthcare resources.

Policy makers in the national 

Ministry of Health and other 

stakeholders and development 

partners should urgently consider 

concrete steps to

increase access to 

modern contraceptives to women 

who need them. 

24

Page 25: Projecting the Cost-Effectiveness of Universal Access to ...

THE END

Thanks Very Much!

25