Improving contraceptive method mix

Jane T. Bertrand, PhD, MBATulane SPHTM

Advance Family Planning Meeting: 9 April 2019

The distribution of contraceptive methods used in a given country reflects:

SUPPLY: Methods that are available and affordable

DEMAND: Consumer preferences, social norms

Two measures of method mix (distribution)

Measure #1: Method skew

• More than 50% of users reply on a single method (Bertrand et al, 2014)

• FP2020 – uses 60% as the criterion

• Skewed method mix is worrisome!

Example: Sudan: 2014 MICS

Pills 78.3%

Causes of a skewed method mix

Acceptable reasons• Community norms favor a particular method• “Everyone in my community uses the injection,

so that’s what I want.”• Historical: Zimbabwe is a “pill country” because

of a strong CBD program years ago

Worrisome reasons• Pressure from a service provider to use a specific

method• Lack of access to a facility with a full range of

methods• Inability to afford the preferred contraceptive• Stockouts of the preferred method

Second measure of method mix: “Average Deviation” (AD)

• A measure of “evenness” of method distribution (Ross et al, 2015)• If users were evenly divided across 8

methods, each method = 12.5%• AD = simple average of differences

around the mean• Potential range of AD: 0-21.9

• 0 if totally “even” (see pie graph)• 21.9 = all users rely on 1 method• Actual range: 6-19• “Middle 50%”: 8.6-12.2

Female Ster.

Vasectomy

IUD

ImplantInjection

Pills

Condoms

Traditional

Improved method mix and increased CPR may go hand in hand (but not always)

Rwanda (1983): CPR = 11.0%

Rhythm

Rwanda (2014/15): CPR = 53.2%

Injectable

Rwanda is the SSA country with the greatest method balance improvement (Ross et al, 2015)

Let’s look at method mix in the 8 AFP countries

• What is the method mix as of the last nationally representative survey?

• Do over 50% of users rely on a single method (“method skew”)?

• What is the “average deviation” (AD) from “evenness”?• The sweet spot ranges from 8.6-12.2

• If skewed, is the country evolving toward a more balanced method mix?

Female ster.7%

Vasectomy2% IUD

1%Implant

3%

Injectable20%

Pill43%

Male condom

10%

Traditional 14%

Method Mix:Bangladesh 2014

CPR: 62%mCPR: 54%Mean AD score: 10.6

AD: Bangladesh 1976-2014

0

2

4

6

8

10

12

14

16

18

20

19761979

19811983

19851988

19891991

19931996

19992004

20072011

20122014

AD V

alue

Survey Year

Indicates skewed method mix

Female ster.3%

Vasectomy0.5% IUD

1% Implant4%

Injectable6%

Pill4%

Male condom

17%Traditional 65%

Method Mix: DRC 2014

CPR: 20%mCPR: 8%Mean AD score: 15.5

AD: DRC 2001-2014

Indicates skewed method mix

0

2

4

6

8

10

12

14

16

18

20

2001 2007 2010 2014

AD V

alue

Survey Year

Female ster.68%

Vasectomy0.6%

IUD3%

Pill8%

Male condom

10%

Traditional 11%

Method Mix: India 2016

CPR: 53%mCPR: 48%Mean AD score: 14.0

AD: India 1981-2016

Indicates skewed method mix

0

2

4

6

8

10

12

14

16

18

20

1981 1988 1993 1999 2003 2006 2008 2016

AD V

alue

Survey Year

Method Mix: Indonesia 2017

Female ster.6%

IUD7%

Implant9%

Injectable52%

Pill20%

Male condom

2%

Traditional4%

CPR: 61%mCPR: 59%Mean AD score: 13.1

AD: Indonesia 1976-2017

0

2

4

6

8

10

12

14

16

18

20

19761979

19851987

19911994

19972003

20052006

20072008

20092010

20112012

20132014

20152016

2017

AD V

alue

Survey YearIndicates skewed method mix

Female ster.5%

IUD6%

Implant30%

Injectable45%

Pill8%

Male condom

3%

Traditional 3%

Method Mix: Kenya 2016

CPR: 61%mCPR: 60%Mean AD score: 10.5

AD: Kenya 1978-2016

0

2

4

6

8

10

12

14

16

18

20

1978 1984 1989 1993 1998 2003 2009 2014 2015 2016

AD V

alue

Survey Year

Indicates skewed method mix

Female ster.2%

IUD5%

Implant11%

Injectable34%Pill

18%

Male condom

9%

Traditional21%

Method Mix: Nigeria 2017

CPR: 13%mCPR: 11%Mean AD score: 10.5

AD: Nigeria 1982-2017

0

2

4

6

8

10

12

14

16

18

20

1982 1990 1994 1999 2003 2007 2008 2011 2012 2013 2016 2017

AD V

alue

Survey YearIndicates skewed method mix

Female ster.9% Vasectomy

0.3%IUD2%

Implant18%

Injectable33%

Pill15%

Male condom

6%

Traditional17%

Method Mix: Tanzania 2016

CPR: 38%mCPR: 32%Mean AD score: 10.3

AD: Tanzania 1992-2016

0

2

4

6

8

10

12

14

16

18

20

1992 1994 1996 1999 2004 2005 2010 2016

AD V

alue

Survey Year

Indicates skewed method mix

Female ster.7% IUD

2%

Implant19%

Injectable47%

Pill7%

Male condom

6%

Traditional12%

Method Mix: Uganda 2017

CPR: 38%mCPR: 34%Mean AD score: 10.0

AD: Uganda 2006-2016

0

2

4

6

8

10

12

14

16

18

20

2006 2011 2014 2015 2016 2017

AD V

alue

Survey Year

Indicates skewed method mix

Three take-away messages

1) The predominant method differs by country: (*) indicates >50% skew• Injectable: Indonesia (*), Kenya, Nigeria, Tanzania, Uganda• Female sterilization: India (*)• Pill: Bangladesh • Traditional: DRC (*)

2) Four patterns emerge from 8 countries:• Method skew has improved but CPR is low (DRC, Nigeria)• Method skew persists but CPR is relatively high (India, Indonesia)• Method mix is balanced with relatively high CPR (Bangladesh, Kenya)• Method mix is balanced but CPR < 35% (Tanzania, Uganda)

3) There is no “ideal method mix,” but method skew raises red flags:• Do clients have access to a full range of accessible and affordable methods, free of

provider bias?

Advocacy has a role in improving method mix

• Increase awareness of method mix among country leadership• Strengthen role of the private sector (e.g., EC, DMPA-SC)• Ensure availability of all methods in public and private sectors• Include all methods on the essential medicine lists, including

pharmaceutical medicine lists• National and subnational governments to budget for contraceptives,

equipment, personnel, and training• Operationalize task sharing guidelines at national and subnational

levels• Address barriers to contraceptive access for young people

Acknowledgements

• Dr. John Ross, independent consultant and expert on method mix

•Ms. Kate Schultz, Tulane graduate student